5 Basic Questions to Ask Before a Pay Equity Study

By Zhuang Liu

Starting a pay equity study from scratch can be a daunting task, and a successful outcome depends on careful preparation. Before diving in, organizations should step back and ask some fundamental questions to clarify the study’s purpose, scope, and methodology to ensure results are meaningful and actionable.

Here are five key questions we recommend asking and answering before launching a pay equity study:

The first step is to define your organization’s goal. Are you conducting the study as an internal audit to better understand pay practices? Or are you conducting the study to gauge compliance with legal or regulatory requirements?

Your objective will shape the framework of your analysis, including which employees to include, what data to collect, and how to structure comparisons. For example, a company-wide study to examine compliance with Title VII of the Civil Rights Act (Title VII) may look different than a pay equity study aligned with The Massachusetts Pay Equity Act (MEPA) or a pay study to evaluate your pay practices (e,g., do merit factors valued by your company actually predict employee pay?).

Make sure that you involve legal counsel, stakeholders, and leadership before undertaking a study.

Data collection takes time, so it is always good to lay out what data your organization needs before the study begins.

• Some pay equity studies begin with total compensation. Given the distinct types of pay that comprise total compensation such as base pay, bonuses, and overtime, you should exercise caution when total compensation is the pay variable under study.
• You will find more value from a pay study that focuses on more distinct pay variables including base salary, bonuses, and/or merit increases which may have different eligibility criteria and may require different comparison groups or different job related factors that influence them.
• More complicated studies may extend to components like equity-based compensation, benefits, etc.
• Other considerations when choosing a pay variable include the need to annualize pay for part-time employees and decide how to handle pay in different currencies when undertaking an international pay study.

Many pay equity studies follow a Title VII framework where employees are grouped into “similarly situated employee groups” (SSEGs). State-based pay analyses may require slightly different employee groupings based on the specific regulatory requirements. In practice, the choice of grouping often reflects a balance between having a bigger sample size to meet statistical thresholds and having better comparability to meet the similarly situated definition:

• Narrower groups (e.g., job title) may ensure the “apple to apple” comparisons but also may lack enough employees per group to apply meaningful statistical testing. For example, regression analysis requires a certain number of employees in an SSEG.
• Broader groups (e.g., job family and grade) may ensure the sample size is large enough to utilize most statistical tools but risk mixing employees performing different work, which can obscure true disparities or uncover false ones.

Finding the right balance is sometimes more of an art than science. In practice, one can always start with something simple like job titles, examine the sample size and statistical coverage of current groupings, and then adjust groupings accordingly.

To carry out a successful pay equity study, you will also need to identify employee characteristics (i.e., merit variables) and structural characteristics (e.g., market data) that influence compensation. The right choice depends on your organization’s pay practices and data availability.

Common variables of interest include:

• Time factors such as time in job, other time in company, age at hire (as a proxy for prior experience)
• Other merit variables such as performance ratings and management level
• Market data such as geographical differentials if your workforce spans regions with different pay norms and market midpoints which capture the median salary for a specific job based on an external salary survey. Market variables can be useful controls in addition to time factors and performance.

Lastly, it is important to think about the statistical tools to employ in your analysis. The choice of method depends on group size, data quality, and study objectives. There are several common statistical tools:

• Regression analysis is ideal for large groups since it provides an estimate of the ‘value’ of different merit factors and can identify demographic group pay differences after controlling for merit factors.
• Fisher’s exact test can be a useful supplemental statistical tool for identifying demographic group pay differences in groups that are not large enough for regression analysis but still have a certain number of employees in each gender or race group.
• Non-statistical comparisons can be used to examine demographic group pay differences for groups that are too small for formal statistical tests. It compares simple averages or medians and identifies any groups with disparities that are alarmingly large.

These questions provide a useful starting point for building your organization’s pay equity study.

If you need more guidance, our team of experienced consultants at DCI is ready to help you design, analyze, and implement a study tailored to your organization’s needs. Visit our Pay Equity page for more information.