New tools promise quick fixes for pay equity. Exercise caution.

October 15, 2019

Mercer’s Pay Equity Calculator, together with our leading experts, can help you get it right!

The promise: Your data loaded in a flash. Complex statistics you can gloss over, run at the push of a button. Clear, immediate solutions.

What’s the reality? Data needs to be relevant and checked. Black-box measures are tough to defend. Pay adjustments should drive progress but also square with your rewards philosophy.

Mercer’s approach is built to address these realities, informed by years of impactful collaboration with the world’s leading companies. We effectively partner with you, by pairing our expertise, on both the data required and statistical methods applied to workforce issues, with yours, on your compensation system and specific employee situations. For analysis, our labor economists and psychologists bring the domain knowledge and applied experience, which is hard to find, build or replace with technology. Our software, the Pay Equity Calculator or PEC, is oriented, not to analysis, but to your consideration of the issues and options to address them. The PEC guides you through an efficient, disciplined review process, which can be consistently applied across a large, global enterprise.

Simply put, we rely on both consulting expertise and unique technology to best partner with companies – and to get it right!

Our approach, depicted here, consists of three steps, each pursued in concert with our clients.

On Data Preparation. We consider possible fields for collection based on academic research, legal precedent, and our experience in the field. We refine the list based on your rewards programs – the factors that should matter and that you would be willing to defend as driving differences in pay – and based on what’s tracked in your systems. While most of our clients then produce the data to a final, agreed-upon specification, we run basic checks to ensure the ultimate integrity of that data.

Loading data into our environment is a fast process, though getting it right requires some due diligence, discussion, and, not infrequently, iteration. 

On Model Development. Mercer’s engagement is heaviest in this step, as this is where the return to deep expertise is realized. While analyses – multiple regressions, to be exact – run in seconds, the work is about getting to the right set of analyses. There are two critical considerations.

1. Workforce segmentation. Our consultants engage with your compensation teams to identify workforce populations subject to difference compensation practices or rules (i.e., how different factors, like employee performance and experience, drive pay). These rules, for example, might be intended to vary by business area, geography (e.g., region or country), or job function, which would then recommend segments for separate statistical models. Segmentation decisions are then tested, with appropriate statistics, to ensure that employees grouped together are, indeed, paid according to the same norms.

Segments for modeling do not define how differences by gender or race/ethnicity will be analyzed, but rather what each individual is expected to be paid, when compared to others who are paid according to the same rules. Testing for differences by gender and race/ethnicity occurs in the Pay Equity Assessment step, generally among “similarly situated” employees according to the legal standard or at the level of a business unit led by a particular manager.

2. Statistical Modeling. In each segment, multiple regression analysis predicts pay from potentially legitimate drivers, which exclude gender and race/ethnicity. Each of the resulting “models” needs to be checked to ensure that it explains sufficient variance in pay – where it does not, additional controls might be considered. In addition, the factors driving pay need to be checked to ensure their effects are aligned with the rewards philosophy, as expected pay rates for employees and, ultimately, pay adjustments will be based upon these models. Review of the estimated models across segments as well as comparison models run for other companies can provide guidance on potential improvements and where caution should be taken in considering the results.

Getting to the right models requires experience evaluating their performance, facilitated open review of the details with company compensation teams and legal counsel, and again, iteration. The work requires bringing together all the experts and, truly, requires adequate time for review and refinement.

On Pay Equity Assessment: This step is where the PEC comes in, and where our client partners often take over, to the extent they have the time and resources. The PEC takes the statistical models as inputs, allowing compensation professionals to investigate and resolve issues- focusing on any area of the company, assessing risk, and effectively responding through targeted pay adjustments. A user also can consider unit-level and company-wide pay gaps, all increasingly under demand by business leaders and activist investors, and various broad remediation options, subject to budget constraints. Finally, a user can consider specific employee cases, and documenting non-standard situations for which there are defensible explanations. The PEC will instantaneously recalculate pay gaps and budgets based on your remediation decisions.

Getting to the right outcome requires pitting actions against identified risks and desired outcomes, with confidence. The PEC is designed for this purpose.

Getting pay equity analysis right means a lot more than quick response. It means genuinely supporting firm, company commitments. Our process, end-to-end, is built to ensure you can engage the right people, at the right time, effectively and expeditiously, to meet those commitments, providing an accurate measure of risk and, where necessary, driving the most efficient remediation.

Brian Levine, Ph.D.
by Brian Levine, Ph.D.

Partner and Pay Equity Leader, Workforce Strategy & Analytics Practice

Dan Lezotte, Ph.D
by Dan Lezotte, Ph.D

Principal, Workforce Strategy & Analytics