We care deeply about equitable compensation (hence our name; Pequity = pay + equity).
We’re building tools and processes to change compensation programs from being reactive (as most currently are) to proactive programs – surfacing insights to compensation decision makers at the time of, or before, a decision is needed.
Does your company have a fair pay program?
We’re here to help you answer that. We’ll explain pay equity tests, and provide some industry best practices to evaluate your programs.
Pay equity tests can be as simple or as complicated as you want them to be, but they are generally based on pay equity law principles. In California this is defined as:
“equal pay for employees who perform ‘substantially similar work,’ when viewed as a composite of skill, effort, and responsibility” (California Equal Pay Act).
Google’s re:Work team has an amazing article on how to run a pay equity test in greater depth, and we’ll walk you through high-level steps of a pay equity test using California law.
Disclaimer: pay equity tests should have legal oversight, as these tests may uncover accidental pay bias that require immediate correction – which can cause strong reactions. A prime example of this is when Google ran a pay equity test and found they were underpaying men in some roles, resulting in raising those salaries – leading to criticism.
Let’s kick-off a simple pay equity test comparing the pay of men and women in software engineering (not considering performance). Pro-tip: You can swap out gender with race/ethnicity to run those types of pay equity tests.
Let’s get started! You’ll need the following data for each employee:
- Gender
- Salary
- Bonus or other cash payments
- Role
- Level
- Location
Notice equity is not part of this check. Most companies, especially early stage, omit equity from pay equity tests (which is legal to do) as equity varies so frequently based on when an employee started, the stock price at time of hire, and company performance since then. As a result, it’s difficult to prove if pay differences are due to gender, race, or just due to timing and market movement.
Now that you have the data needed to run a pay equity test, next step is to check the distribution for men’s pay, and the distribution for women’s pay for each of these groups using a regression model (you can do this in Excel or Google sheets). The industry standard way to evaluate your test is if the mean pay of one gender is 2 standard deviations above or below the other gender’s mean pay, you have a pay disparity. If the pay is 1 standard deviation away, it may not be a disparity but it should still be reviewed for better understanding.
For smaller companies, there may not be enough women or underrepresented minorities to have a significant n-count to test. In these cases, a faster, simpler way to do this test is to check if women’s pay is >5% from their similarly leveled & performing male peers, in the same role. If the answer is yes, then you should dive into their pay differences and consider correcting to be <5% away from their peers.
Both methods above are quick and dirty ways to approach pay equity testing, but are usually enough to get companies started.