In 2018, a UK law came into force that all companies with over 250 employees must report their gender pay gap. The results are instructive, principally because they make concrete something we already knew about but could never quite pin down: It’s one thing to understand that your country or industry pays women, on average, less than men, but the knowledge has a vastly different power when it becomes company-specific.
Helen Matthews, chief people officer at Ogilvy UK, one of the country’s biggest advertising agencies, tells Quartz that reporting pay-gap figures (pdf) has turned an internal spotlight on the issue. It’s now regularly discussed at the company.
But a massive number of employers are missing out on the opportunity to address their pay-gap problem because they simply haven’t tried to calculate it, and don’t know how to, says Andrew Chamberlain, chief economist at Glassdoor, which allows employees to self-report salaries and other information about their workplaces. Glassdoor’s 2019 report (pdf) analyzed reported data and found that on average women in the UK were paid 17.9% less than men. In the US they earned 21.4% less.
“I’m starting to think the biggest opportunity for companies to do something about this today, and improve [the gender pay gap], is to dig into their DNA, inside their HR organization, to start studying this on an ongoing basis and being transparent,” Chamberlain said.
“It’s really easy to do with your own payroll data. We do it at Glassdoor. And we have a guide for employers that gives them software…that lets them do this themselves.”
This guide is a 12-page document that walks anyone through the necessary steps for measuring their gender pay gap. It explains how to calculate both a “raw” gap between average male and female earnings, and an “adjusted” gap that takes into account factors like seniority and educational attainment, creating a better like-for-like comparison. The guide is practical, including sections on how to obtain, organize, clean, and protect employee data, and downloadable code (using the R language) with which to run the analysis.
The guide is not without limitations. Because it’s designed to work on larger datasets, Glassdoor suggests the advice isn’t suitable for companies with fewer than 200 employees. For smaller companies, Glassdoor says, “a case-by-case analysis of gender pay difference by job title may be a more appropriate way to perform a gender pay audit.” And it does require some technical know-how. Though the guide is clearly-written and designed to be as user-friendly as possible, it also calls for at least some understanding of regression analysis and how to work with data.
“I’m quite optimistic, if we could just get more technical people in HR who use their payroll data and study it and share it, they could potentially see benefits themselves as a company, reducing legal risk and giving them something to brag about when they hire,” Chamberlain says. “But they could also eliminate these gaps overnight.”