The “raw” gender pay gap is 78.06%, and the “normalized” gap is 95.24% according to Pave data. Which gender gap benchmark should you be using? _____________ 𝗠𝗲𝘁𝗵𝗼𝗱𝗼𝗹𝗼𝗴𝘆 𝗔: 𝘁𝗵𝗲 “𝗿𝗮𝘄” 𝗴𝗲𝗻𝗱𝗲𝗿 𝗽𝗮𝘆 𝗴𝗮𝗽 The “raw” gender pay gap is the benchmark you are likely already familiar with. When you read about the gender pay gap for salaries in the news, it almost always refers to this “raw” pay gap. In other words, take the average salary for employees who identify as women in a population and compare it to men in that same population. This “raw” pay gap is also the basis of Europe’s DEI reporting requirements. 🚨 “Raw” gender salary gap in Pave’s dataset for USA employees => 78.06% In other words, on average, employees in Pave’s dataset identifying as women make 21.94% less than the employees identifying as men. _____________ 𝗠𝗲𝘁𝗵𝗼𝗱𝗼𝗹𝗼𝗴𝘆 𝗕: 𝘁𝗵𝗲 “𝗻𝗼𝗿𝗺𝗮𝗹𝗶𝘇𝗲𝗱” 𝗴𝗲𝗻𝗱𝗲𝗿 𝗽𝗮𝘆 𝗴𝗮𝗽 Our team dug one level deeper and looked at the gender pay gap normalized by level and job family. Simply put, what does the median women-identifying employee make requisite with her male counterpart in the same job family, job level, and country (USA). 🚨 “Normalized” gender salary gap in Pave’s dataset for USA employees => 95.24% In other words, if you compare two employees with the same job family and job level, there will be, on average, a 4.74% difference in salary between the men and women. 4.74% is certainly not 0%, but it is a lot smaller than the 21.94% delta in the “raw” gender salary gap benchmark mentioned above. 𝗧𝗵𝗲 𝗯𝗶𝗴𝗴𝗲𝘀𝘁 𝘂𝗻𝗱𝗲𝗿𝗹𝘆𝗶𝗻𝗴 𝗶𝘀𝘀𝘂𝗲 𝗶𝘀 𝗿𝗲𝗽𝗿𝗲𝘀𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻 𝗮𝘁 𝗵𝗶𝗴𝗵𝗲𝗿 𝗽𝗮𝘆𝗶𝗻𝗴 𝗹𝗲𝘃𝗲𝗹𝘀/𝗳𝘂𝗻𝗰𝘁𝗶𝗼𝗻𝘀 𝗿𝗮𝘁𝗵𝗲𝗿 𝘁𝗵𝗮𝗻 𝗱𝗶𝘀𝗽𝗮𝗿𝗶𝘁𝘆 𝗮𝘁 𝘁𝗵𝗲 𝘀𝗮𝗺𝗲 𝗹𝗲𝘃𝗲𝗹/𝗳𝘂𝗻𝗰𝘁𝗶𝗼𝗻. _____________ Takeaways: 1️⃣ The massive (~5x) delta between the “raw” and “normalized” gender pay gap is largely driven by much higher representation of men in higher levels and higher compensated job families. For instance, only 20% of C-level employees (E9) in Pave’s dataset identify as women per the second attached chart. 2️⃣ European DEI reporting requirements center around the “raw” rather than the “normalized” gender pay gap. There are pros and cons to Europe’s emphasis on the “raw” rather than “normalized” pay gap. For instance, we hosted 11 customers at our office yesterday and talked about how Europe's emphasis on the “raw” gender pay gap perhaps oversimplifies the matter. That said, a European regulator likely responds “okay, well then just hire more women CXOs” to the company who complains about the oversimplification. My two cents–aside from what the reporting requirements are region-by-region, I would argue that both gender gap measurements–”raw” and “normalized”--are important measures to keep in mind. Go deep and understand the “why” behind both figures and use it to inform your decision marking. #pave #genderpaygap #benchmarks
Happy that the team decided to look into this in detail (I remember us frequently wondering about it, but prioritizing other work). Super interesting results, thanks so much for sharing!
This analysis is very important to consider. The overall comparison of male salaries versus female salaries does not take into consideration the jobs being performed. We should always compare M vs F in comparable roles and level of work to obtain an accurate variance. In some situations woman have consciously made the decision to not take high level management roles in order to be more available for family priorities.
I think the normalized measure works in many scenarios, but the I can see why Europe wants to use raw gender gap. This article is few years old, but I found it super interesting https://www.payscale.com/career-advice/when-an-occupation-becomes-female-dominated-pay-declines/
Thanks for covering this topic Matt Schulman. I find that “gender pay gap” is one of the most misunderstood terms in comp. Thanks for breaking it down in such a clear way. The raw vs normalized explanation is typically the missing piece for folks.
All I read is that the glass ceiling is still alive.
Both!
Hi Matt, very clear the impact of using level and family to dig deeper in the gender pay gap. And if you add the third variable 'time in position', the % difference will go even smaller!

Very insightful. In your example, pay equity in itself is not the issue. The issue as you stated is in available talent among women for the higher paying roles, recruiting and promotional practices employed. I emphasize "available talent" due to the examining the number of women applicants for the higher level roles, mentorship in readying interested internal women for the higher level roles, and outreach by the company to get the company and its opportunities on the radars of talented women. It is less of a compensation issue in the scenario presented than a host of other potential issues outside of compensation Compensation can't always solve everything.