UMD Study: To Close Gender Pay Gap, Use Analytics

While companies often raise the salaries of the most underpaid women to close the gender pay gap, a recent study by a University of Maryland business researcher and colleagues fleshed out two reasons why this doesn’t work.

Margrét Bjarnadóttir, associate professor of management science and statistics in the Robert H. Smith School of Business, co-authored the analysis that found inequity isn’t usually evenly spread throughout the organization, and that the number of high-earning men far outstrips that of high-earning women.

“If you’re always focusing on negative outliers and people who are really underpaid, you’re not addressing this bias in your pay structure, which is often the lack of well-paid women,” she said. As a result, even if adjustments are made to bring salaries more into line, women are stuck around the median, while more men earn lofty pay.

Nationwide, women earn 82 cents for every dollar earned by men, a stubborn gap that has hardly budged in two decades, according to a recent analysis. Bjarnadóttir and co-authors David Anderson of the Villanova University School of Business and David Ross of the University of Florida’s Warrington School of Business Administration explored methods to tackle the problem in the paper published in the journal Production and Operations Management.

Bjarnadóttir and Anderson are cofounders of PayAnalytics, a company that makes compensation analytics software to help firms monitor and address demographic pay gaps, while Ross serves as an advisor to the company, and the publication contains a case study of their work with a large firm.

The pair designed an algorithm that first measures whether an employer has a pay gap and then suggests salary adjustments to close it. “What we do is study how pay is determined,” whether it’s experience, achievements or hierarchical position, she said. “Then we study whether these factors that drive pay account differently for men and women or white and Black people or any other demographic groups. When you find there’s a difference, that’s what we correct.”

PayAnalytics’ suggestions aim to correct pay structure biases in pockets of a company where pay is most unequal, and target employees who are facing inequities for raises. The targeted approach doesn’t lead to wage compression, which refers to employees earning similar pay regardless of differences in knowledge, skills, experience, etc. “We don’t focus on whether or not somebody is paid less than what we expect, but on who is facing the most bias and those are the employees that we target for raises,” Bjarnadóttir said.

Increasingly, companies are taking a proactive approach to pay equity, making it part of their day-to-day operations by incorporating it directly into compensation decision processes. Still, achieving equal pay for equal work can take time.

“Often when organizations are studying their pay equity for the first time, they have what I call a legacy gap: They’ve been making these decisions for a long time, and pay inequities have built up,” Bjarnadóttir said. “Typically, there’s a need for an initial investment that is larger than what it will be in the future, but you need to invest in closing these pay inequities.”