"This paper examines the effects of grid compression on gender-based salary gaps. In this case, the workplace setting is a large Canadian research university previously found to have both a positionally segregated cohort of academic staff and a persistent gender-based differential in their salaries. The annual salary grids at this institution underwent seemingly subtle compressions over the course of two decades, with within-rank compression largely confined to the first decade and across-rank compression confined to the second decade.

We employ a simulation methodology to check whether the two types of compression reduce the salary gap to varying degrees. After deflating staff salaries back to the start of each decade, we project the salaries forward using the historical annual increases and grids in place in each year at the university. We compute the gap present in the salary distribution at the start and end of each decade in the simulation, and check whether the gap decreased more across one decade than the other.

We find that across-rank compression of the institution's salary grid during the second decade narrows the salary gap to a greater extent than the within-rank compression of the earlier decade. Our work demonstrates that employers using stated salary grids could use simulation to monitor the equity effects of their pay policies and shows that they could accelerate the closure of gaps through consciously altering relationships among pay levels at different points in the grid's hierarchy."

Support: Digital