There’s been a lot of discussion in past days about OFCCP’s new Compensation Directive, which describes the Agency’s approach to investigating pay discrimination. We’ve spent some time digging into the details of the directive. We’ve previously shared our thoughts about the Directive’s commitment to transparency, and now, as promised, want to share our thoughts about the nuts and bolts of the Directive. In summary, while the Directive is certainly transparent and descriptive and contains clarifications, it is by no means the “windfall” for contractors as some may be reporting.

The Directive breaks down, in much appreciated detail, the Agency’s approach to (1) creation of pay groupings for analysis, including insights into how the Agency defines “similarly situated” and, (2) statistical modeling.

Pay Analysis Groups (PAGs) are Here to Stay

The contractor community was first introduced to PAGs when OFCCP released the now rescinded Directive 307 in 2013. Since that time contractors have struggled to understand, and predict, how OFCCP groups their workforces for pay analyses. In its new Directive, the Agency takes steps to provide insight into its grouping philosophy. First, the Agency states directly that it “aligns its compensation compliance evaluation procedures with principles under Title VII.” Specifically, with respect to groupings, the Directive states OFCCP defines similarly-situated employees as those who would be expected to be paid the same based on:

  • job similarity (e.g., tasks performed, skills required, effort, responsibility, working conditions and complexity); and
  • other objective factors such as minimum qualifications or certifications.

While the Directive states it is

OFCCP’s objective to use PAGs that mirror a contractor’s compensation system,

the Directive also provides OFCCP the latitude to disregard the contractor’s structure if (1) OFCCP determines the structure is not “reasonable,” (2) cannot “verify the structure as reflected in the contractor compensation policies” and (3) “the analytical groupings are of insufficient size to “conduct a meaningful system statistical analysis.” In an important footnote to the Directive OFCCP expands on these caveats (and its latitude) to say

[t]here is no bright-line rule, and professionals may disagree, about what constitutes sufficient size for a meaningful systemic statistical analysis of an employer’s compensation system. In its preliminary analysis, OFCCP generally seeks to evaluate groups containing employees under a similar pay system performing broadly similar job functions, regardless of group size.

(emphasis added). The Directive further clarifies that OFCCP may “broaden or narrow” its preliminary PAGs based on information received during the review. In the absence of information about a contractor’s compensation system, the Directive states “OFCCP will conduct its preliminary desk audit using either EEO-1 or AAP job groups” where OFCCP will control for “sub-groupings, functions, units or titles, as appropriate. Thus, while it’s goal is to use mirror contractor compensation systems, the Directive certainly contemplates, as a default, OFCCP developing large, and broad, groupings for analysis.

OFCCP’s Statistical Methodology and Modeling

After discussing the development of pay groups, the Directive moves on to detail the principles OFCCP will use to analyze the groupings. Utilizing technical terminology, OFCCP lays out how it will analyze pay, the variables it may use to build its regression model for the purposes of analyzing pay, and how it may treat those variables in the model. The principles the Agency will apply include:

  • Separately analyzing base and total compensation
  • Transforming salary to the natural log (which will help account for different pay distributions within a pay group)
  • Controlling for time variables separately (e.g. analyzing time in job separate from time in company)
  • Considering squaring tenure variables (which will allow the model to take into consideration the natural growth, and eventual stagnation, of earning potential)
  • Considering education and performance ratings
  • Considering combining categories of variables to better capture pay differentials (e.g. analyzing the impact of PhD and Master degree requirements together versus lesser education requirement instead of analyzing the impact of each degree separately)

Again, while the transparency is welcomed, and overdue, the Directive does not require the Agency to utilize the principles or build its model in a specific way. Thus, the Agency will continue to have the latitude and autonomy to develop its methodology on a case-by-case basis. What remains to be seen is how much input OFCCP will allow contractors in this process and how much transparency the Agency will give into the methods and modeling it concludes are appropriate.