When performance trumps gender bias: joint versus separate evaluation

Bohnet, I., Van Geen, A., & Bazerman, M. 2016. Management Science 62(5): 1225-1234

The authors of this study examine a new intervention designed to reduce the impact of gender stereotypes on hiring and promotion decisions, by evaluating people’s future performance jointly rather than separately. The paper shows that joint evaluation allows people to be selected based on past performance rather than irrelevant characteristics such as gender. 

SUMMARY

Recent research shows that learning the sex of a person during an evaluation leads to gender biases and “implicit discrimination that is not based on a rational assessment” of performance. To overcome this gender bias in hiring and promotion, the paper examines a new intervention, designed to evaluate people in a group rather than separately, what the authors call an “evaluation nudge”. The hypothesis is that when evaluating multiple candidates at a time, evaluators will be less influenced by gender stereotypes and will switch to a more comparative, information-based approach. The paper shows that this joint-evaluation is able to help employers take into account performance, rather than gender stereotypes. 

Methodology 

They experimentally examine whether the intervention has an impact on the decrease of gender gaps in hiring, promotion, and job assignments. The experiment is designed to simulate an evaluation process where employers decide the candidate’s suitability for a job.  They conducted the experiment with a total of 654 individuals, where 554 played the role of employer and 100 played the role of employee, with an equal number of male and female employees. To study the impact of gender stereotypes, they used “two sex-typed tasks, a math and a verbal task”, and asked employers to choose an employee based on his performance in these tasks. 

Findings

The findings show that “gender was not correlated” with future performance on the tasks, but individual performance was predictive of future performance. But in separate evaluation, employers were more likely to be influenced by gender than by past performance. In contrast, “in joint evaluation, gender was irrelevant” and past performance plays a more important role; “employers were more likely to choose the higher-performing employee”. “51 percent of the employers” in separate evaluation “chose the under-performing employee”, in contrast with only 5 percent of the employers in joint evaluation. 

This change in decision making based on the evaluation mode is compatible with the research in behavioural decision making, which suggests that the informational component in joint evaluations provides more data to the employers and makes them rely less on intuitive stereotypical beliefs and more on a reasoned approach. 

These findings have practical implications on evaluation procedures, for the organisations that want to decrease the impact of gender stereotypes in hiring and promotion decisions. “Organisations may be able to nudge evaluators toward taking individual performance information into account rather than gender stereotypes by introducing joint rather than separate evaluation procedures”.