Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This paper proposes a new decomposition method to understand how gender pay gaps arise within firms. The method accounts for pipeline effects, non-stationary environments, and dynamic interactions between pay gap components. The decomposition is applied to a new data set covering all employees at the World Bank Group between 1987 and 2015 and shows that historical differences in the positions for which men and women were hired account for 77% of today’s average salary difference, dwarfing the roles of entry salaries, salary growth, or retention differences. Forward simulations show that 20% of the total gap can be assigned to pipeline effects that would resolve mechanically with time.