Even though pay differences between men and women have declined in past decades, evidence suggests men continue to receive greater pay raises than women for the same performance.
In a recent study, I investigate whether these biases in annual merit raises disappear when supervisors get to know new employees over time.
I found that no matter how long employees work with their company, gender pay gaps widen with each additional year employees stay.
This means that gaps in annual merit increases repeat every year, causing overall gender pay gaps to widen a little more with each year of tenure. This pattern happens even when employees have the same education, work in the same job, and receive the same annual performance ratings.
But, there is good news. HR interventions are very effective in limiting these trends. Policies prevented gender pay differences at hire altogether and in some instances, even limited subsequent gender gaps.
My results are based on the analysis of over 20,000 de-identified HR records from a private U.S. employer between 2005 and 2014. These data allow me to examine how earnings change after employees are hired. I can also account for factors that typically explain gender pay differences, such as annual performance ratings or job type.
What do widening gaps after hire tell us about the nature of inequality?
Social scientists have long debated why some supervisors reward men more than women.
One perspective emphasizes managers know little about applicants when hiring external candidates. As a consequence, managers draw on their own assumptions related to race or gender to estimate applicant’s future productivity. These assumptions can determine how much pay managers offer candidates.
However, once managers get to know their employees, they replace assumptions with observations and adjust pay to employees’ actual performance. As a consequence, gender pay gaps would narrow among equally performing employees after hire.
Another perspective argues that beliefs about men and women operate like a mental lens through which supervisors filter all information.
Research revealed that while most individuals are unaware, they often perceive women as less worthy of rewards than comparable men. This belief may affect supervisor assignment of performance rewards in their team. If decision makers perceive women’s work as less valuable, then supervisors might give smaller rewards to women, regardless of how well supervisors know these employees.
As a consequence, gender pay differences among comparable employees would increase constantly over time.
My data supports the second perspective. This means that biases are not merely a problem of missing information, but an ongoing mental process that persists even when we have detailed performance records at our disposal.
What can employers do?
Employer practices can both cause and remedy workplace inequality.
The organization in my study went to great length to prevent pay inequality.
It particularly emphasized pay equity at hire. Before managers could extend job offers, compensation specialists used applicants’ qualifications and job responsibilities to determine a salary range. Hiring managers and applicants had to stay within this range during salary negotiations.
My analyses show these efforts succeeded. Men and women were hired at the same salary, once I control for education, prior work experience, pay grade, and occupation.
While the organization focused intently on hiring salaries, less attention went towards later pay decisions. De facto procedures for subsequent merit raises varied substantially between business units within the organization.
In some units, a committee or HR representative decided how to spread rewards among employees. This process effectively limited supervisors’ direct influence over performance-rewards. In other units, supervisors had more discretion and could determine rewards themselves.
A comparison between units showed that gender gaps were absent in units that used committees or HR representatives to determine annual raises: Here, comparable men and women received similar pay for the same performance.
In contrast, men and women’s pay grew apart over time in units where supervisors had more discretion over merit increases. Here, women and men earned the same at hire, but the earnings gap grew more unequal the longer they stayed. For instance, after having worked in high discretion units for eight years, professional women lost on average $9,000 in total earnings.
Together results show that HR interventions can reduce workplace inequality. But, eliminating differences at hire is not enough. Instead, policies need to address later changes in pay.
To prevent disparities, effective policies don’t have to take away decision power from supervisors. Instead, increasing supervisors’ accountability is a better strategy to avoid disparities among similar employees.
Organizations can hold supervisors accountable by regularly collecting data and by asking supervisors to explain their decisions. In a prior study, regular reviews by a performance-reward committee quickly eliminated race and gender differences in annual merit increases. These review committees are most effective when employees from all over the organization (not just HR professionals) serve on them.
Anne-Kathrin Kronberg. “Workplace Gender Pay Gaps: Does Gender Matter Less the Longer Employees Stay?” in Work and Occupations 2020.
Image: GDJ via Pixaby (CC BY 2.0)