Unpacking Gender Bias in Hiring: Insights from Japan's Synthetic Personality Inventory

Sociologists, economists, and psychologists consistently emphasize that an individual's socioeconomic status is influenced not only by cognitive abilities but also by non-cognitive skills, including personality traits (for a review, see Farkas, 2003). Furthermore, it has been repeatedly noted that the impacts of cognitive abilities and non-cognitive aspects often reflect gender bias, contributing to gender inequality in the labor market (e.g., Quadlin, 2019). To comprehensively assess how gender shapes evaluations of cognitive and non-cognitive skills, greater attention should be directed toward demand-side mechanisms.

This paper delves into employers' gender bias regarding job seekers' cognitive and non-cognitive abilities in the Japanese labor market. Japan has widely adopted the Synthetic Personality Inventory (SPI) due to advancements in psychological testing theory. Approximately 80% of private companies in Japan require candidates to take the SPI, which includes basic ability tests and personality assessments when submitting their resumes. Japanese employers can utilize information on candidates' basic ability scores and personality traits based on the Big Five, in addition to resume details such as educational qualifications. This unique job hiring process in Japan allows for the objective verification of employers' gender bias in evaluating applicants' cognitive and non-cognitive skills, as well as the assessment of the relative importance of these traits.

I conducted a vignette study simulating a hiring process for university graduates. I asked 2,095 employers in private companies located in the Tokyo metropolitan area to evaluate 16 sets of vignettes featuring job applicants competing for the same position. Participants were asked to assess the hiring potential based on simulated screens of resumes and SPI test results. The analysis revealed that the gender bias identified in existing research from Western countries was not significantly observed. These findings in a non-Western context suggest that gender bias in ability evaluations is embedded in social contexts.