Dynamics of the Gender Gap in the Workplace: An econometric case study of a large Japanese firm

This paper provides new evidence on the nature and causes of the gender pay gap using confidential personnel records from a large Japanese manufacturing firm. Controlling only for the human capital variables that are typically included in the standard wage function results in a substantial gender pay gap—16% for unmarried workers and 31% for married ones. However, additionally controlling for job level, skill grade, hours worked, and number of dependents almost eliminates the "unexplained" gender pay gap. We estimate various models of promotion rates and additionally find that (i) there is a statistically and economically significant correlation between the hours worked and the odds of promotion for women but not for men; (ii) maternity carries a substantial career penalty (up to a 20-30 percentage-point fall in future earnings), especially for college graduate women; and (iii) the maternity penalty can be avoided by promptly returning from parental leave and not reducing work hours after returning. As such, our evidence points to the importance of women's ability to signal their commitment to work (or the level of family support they receive)—through working long hours and taking shorter parental leave—for their career advancement.