Ever since Massachusetts Governor Charlie Baker signed the state Equal Pay Act (MEPA) on August 1, 2016, employers have been seeking direction on how employee pay should be analyzed to withstand scrutiny under the new law. MEPA goes into effect on July 1, 2018.
On March 1, 2018, the Massachusetts Office of the Attorney General (OAG) issued a Guidance containing explanations and examples for applying the new legal standards, frequently asked questions, tips for conducting a proactive self-evaluation, and a “Pay Calculation Tool” to assist employers with identifying and evaluating gender-based pay gaps under MEPA. Jackson Lewis previously published a detailed discussion of the OAG Guidance. We asked our statisticians to examine the Pay Calculation Tool to better understand how it works and what additional insights it might provide on how to identify and address pay discrimination under MEPA. This article provides a summary of our observations.
The bottom line is that the Tool can help organize relevant data and be used to conduct basic self-evaluations. As the OAG Guidance makes clear, however, the Tool is not appropriate for larger pay groups or sophisticated pay systems. Therefore, most employers should not rely on the Tool alone to meet their obligations under MEPA. Moreover, the calculations produced by the Tool may be misleading or incomplete. Thus, employers should be cautious and obtain legal counsel in evaluating whether the Tool provides the right approach for their organizations.
MEPA broadens the definition of who can be compared for pay purposes in Massachusetts, limits the considerations employers can use to explain pay differences between sexes, and prohibits employers from asking applicants about salary history. The law’s new standards have yet to be interpreted by a court through litigation. Accordingly, there are still many unknowns under MEPA.
One of those unknowns is how robust a “self-evaluation” must be to provide employers the benefit of MEPA’s “safe harbor” provision. MEPA provides an affirmative defense for any employer that has conducted, within the previous three years and before an action has been filed, a good faith, reasonable self-evaluation of its pay practices and taken steps to correct any impermissible disparities.
Though the OAG Guidance does not have the legal force of a regulation, it offers insight into what makes a self-evaluation “reasonable.” The Guidance states that a self-evaluation must be reasonable in “detail” and “scope” in view of the employer’s unique pay system. Further, what works for one pay grouping may not necessarily work for all pay groupings.
In the Guidance, the OAG explains, “For an organization with small, clearly defined groupings of comparable jobs and relatively simple pay structures, a simple analysis comparing the average wages earned by men and women in comparable jobs may be sufficient to identify where there are disparities.” (Appendix A: Self-Evaluations – A Basic Guide for Employers.) To assist employers with this type of simple analysis, the OAG published the “Pay Calculation Tool.”
The Tool is an Excel spreadsheet containing fields for employers to enter employee-specific pay information, assign pay groupings, and quickly see each employee’s base salary and total compensation in relation to others in the same grouping. The Tool also uses embedded formulas to automatically produce helpful metrics, such as the difference in average pay between sexes and each employee’s “gap” from the group average. These calculations can assist in identifying specific pay “outliers” for deeper review or equity adjustments.
The Tool also produces a “Further Review” worksheet with basic statistics on each employee’s pay situation for additional investigation by the employer. This report has open fields for the employer to insert potential explanatory factors (e.g., years of service, years of experience, performance score, location, and so on). This feature apparently is designed to help the employer manually evaluate whether employees’ backgrounds support any gender-based pay differences.
The Tool’s reports can be difficult to interpret, the embedded formulas can produce quirky calculations, and even minor adjustments to input numbers can lead to significantly different results. Moreover, the OAG has cautioned that the Tool is not appropriate for many (if not most) employers’ pay systems. Therefore, employers need to take care, first, in evaluating whether to use the Tool and, second, in how to interpret the output reports.
1. The Tool is Not Appropriate for Use With Large Groups or Sophisticated Pay Systems
The OAG Guidance makes clear that the complexity and type of analysis required to identify discrimination under MEPA “will vary significantly depending on the size, make-up, and resources of each employer.” It also expressly provides that the Pay Calculation Tool is intended only for simple analyses of small pay groups with relatively simple pay structures.
Before using the Tool, the Guidance recommends that “[e]mployers  consult with legal counsel about their options and what type of analysis is most appropriate for their organizations.”
Indeed, the Tool probably is not appropriate for most employers’ pay systems. For employers with larger groupings of employees doing “comparable work” (at least 30 employees), employers with unclear or uncertain pay groupings of comparable jobs, and employers with sophisticated pay systems, the Tool likely is not enough. More robust, statistical analyses are needed in these situations. The Guidance provides:
While not necessarily required by the statute, in many cases, conducting a statistical analysis will be the best way for employers to determine whether there are differences in pay between men and women in comparable jobs after controlling for other factors. In most cases, conducting such a statistical analysis will constitute good faith with respect to this step of any employer’s self-evaluation of its pay practices.
In other words, the Tool does not provide the type of statistical analysis that the Guidance says is likely necessary for most employers to identify and address pay discrimination under MEPA or to obtain the benefit of the self-evaluation affirmative defense. Employers with larger pay groupings or sophisticated pay systems should consider conducting multiple regression analyses, which can pinpoint the extent to which pay differences are explained by job-related data.
Many of the Tool’s embedded “calculator” functions are based on a group median, rather than mean. The “mean” and “median” are essentially two different ways of looking at a group average. The mean is the number resulting from adding a set of quantities, then dividing the total sum by the number of quantities in the set. The median, however, is the number in the middle of the set, if sorted from highest to lowest value. If the set of numbers was 1, 2, and 6, the mean would be 3 ((1 + 2 + 6) ÷ 3), but the median would be 2 (the number in the middle).
In relying on the median, rather than the mean, the Tool easily can create misleading results. For example:
- In the chart below, the median pay for the group is $40,000. Although the male is paid $20,000 more than either female, neither female is below the median pay amount for the group. The “% of Women Below Median Hourly Rate” calculation (on the “Calculation Tool” tab of the spreadsheet) shows “0,” which appears to indicate females are paid fairly in the group.
|Gender||Annual Base Salary|
- However, if we increase one female’s pay by just $1, the “% of Women Below Median Hourly Rate” calculation shows “50%,” which appears to indicate a big problem.
|Gender||Annual Base Salary|
- In addition, the use of medians can hide significant pay outliers that should be examined or mask significant differences in pay amounts. For example, the “% of Women Below Median Hourly Rate” result is exactly the same in the following two, very different scenarios:
|Gender||Annual Base Salary|
|Gender||Annual Base Salary|
The Tool provides calculations and comparisons based on base salary/hourly rates and “total compensation,” but prompts manual review of specific employee situations based primarily on total compensation (see “Worksheet – Further Review” tab of the Tool). This emphasis on total compensation can be misleading, because it generally requires an aggregation of different pay types, earned based on different criteria that potentially include personal choice, personal predicaments, and personal work effort.
In many situations, different pay types (e.g., bonuses, incentives, overtime, and so on) must be evaluated in standalone analyses to provide an “apples-to-apples” view across comparable employees. For example, assume a male employee was paid $60,000 in base salary but was a poor performer and earned a $1,000 annual bonus. Meanwhile, a woman doing comparable work was paid $50,000 in base salary and, because she put in significant extra hours and did additional work, she earned a $9,000 annual bonus. Under a total compensation view, there was only a $2,000 pay gap between these two employees, despite the considerable extra efforts expended by the female. What is troublesome about this approach is that the female’s extra efforts and the resulting large bonus actually may serve to mask the underlying base salary pay gap, which may be impermissible under MEPA.
Statistically, the factors that influence and predict different types of pay vary significantly, thus, an aggregated analysis of multiple pay types typically is not appropriate and may produce unsound results. Practically, if an analysis of total compensation resulted in the identification of a pay gap that needs to be adjusted, it can be difficult to identify which pay component should be adjusted. If an adjustment is made to base salaries with the intention of “fixing” a total compensation problem, what impact does that have on the comparison of base salaries alone? If there is an impermissible disparity in base salaries of comparable employees, would it be excused under MEPA if the disparity disappears once performance bonuses are added? Can the problem even be fixed through pay equity adjustments, or should the employer consider how such things as bonuses, incentives, and overtime are administered?
These questions can be difficult to answer and highlight complications that can arise when trying to analyze total compensation. Therefore, employers should keep in mind that the Tool provides just one view of pay differences that may need to be examined through multiple lenses.
The instructions for the Tool surprisingly direct that all commission payments be excluded from the self-evaluation Tool. The instructions caution employers to review commission formulas for consistency but are silent on how to determine whether actual earnings under such formulas are consistent with MEPA.
While employers should ensure commissions are calculated in the same manner for employees doing “comparable work,” a variety of other factors may affect whether commission-based workers are paid fairly, including territory assignments, quotas, and sales targets. Indeed, whether the administration of such factors is creating pay discrimination should be a critical question in any MEPA analysis.
Analyzing only base salaries of commission-based employees may provide an incomplete picture of a pay system. Employers should take this into account when attempting to use the Tool for employees who are eligible to earn commissions.
5. The Value of the Tool’s “Further Review” Report is Questionable as the Component Prompts Review of All Employees, Regardless of Pay Gap Calculations
The third tab of the Tool workbook (“Worksheet – Further Review”) provides a summary of each employee’s data based on the employer’s manual inputs. While this report provides a convenient way to organize relevant data for a manual review, it does not do much to identify which pay gaps need attention under MEPA. The Tool also does not help employers answer the difficult question of whether job-related background information (e.g., tenure, experience, education, and so on) is enough to justify pay differences.
In addition, every single employee entered into the Tool on the first tab of the workbook is included in the Further Review report on the third tab. The report does nothing to help whittle down employees for further investigation, such as by excluding (1) single person groupings where there are no comparators, (2) groupings in which everyone is the same gender, or (3) groups with little or no “pay gaps” under MEPA. This appears to go beyond MEPA by implying that employers should evaluate everyone under the expanded standards of MEPA, regardless of whether exposure under the new law is even possible.
The Further Review report also can be misleading because it does not include or contemplate all potentially relevant considerations underlying pay differences. The report identifies certain categories of information that may help explain pay differences, but it excludes others, such as business unit, product line, market, budget, and the like. Employers who rely on such an incomplete picture run the risk of making decisions or adjustments that actually may create discrimination.
Unless protected by the attorney-client privilege, the reports created using the Tool likely would be discoverable in litigation and government investigations, including investigations by the Equal Employment Opportunity Commission (EEOC) and the Office of Federal Contract Compliance Programs (OFCCP), which do not recognize the affirmative defense under MEPA. The reports generated by the Tool will identify pay differences that may not be meaningful, yet appear to be a problem in the Tool’s reports. Regardless of how the Tool reports are used, the existence of a report showing meaningful pay gaps can be held against the employer if the employer did not do enough to close those gaps.
Even early drafts of the Tool spreadsheets could be risky if they reveal early thoughts on which employees may be doing “comparable work” under MEPA. This will no doubt be a key issue in litigation under the new law. Plaintiffs’ attorneys also may seek copies of these reports to try to buttress other types of discrimination or favoritism claims.
Therefore, employers should be cautious and consult with counsel before attempting to use the Pay Calculation Tool (or conducting any other type of self-evaluation).
The Pay Calculation Tool provides one option for Massachusetts employers to conduct self-evaluations of small pay groups with simple pay structures to obtain the benefit of the affirmative defense under MEPA. However, the Tool is not appropriate for use with larger groups under more complex pay systems and may produce odd or confusing reports.
As the OAG makes clear in the Guidance, employers should seek legal counsel on what type of analysis is appropriate in view of the size and complexity of pay groupings, as well as the nuances of the company’s unique pay practices.
Employers need to keep in mind that pay is complex and that there is no “one size fits all” approach. Multiple approaches or “cuts” of analyses may be necessary to fully understand whether pay gaps are meaningful or otherwise create exposure for the company under MEPA.