OFCCP Affirmative Action Plan Compensation Analysis
Over the past several years, the OFCCP has increased its focus on detecting pay discrimination in compensation systems affecting women and minorities. Past efforts by the agency to prosecute contractors who they believed had discriminatory pay practices were largely unsuccessful due to flawed detection methodologies. The OFCCP has now regrouped and rearmed with a new Division of Statistical Analysis, which will analyze contractors’ compensation data using rigorous statistical analysis that they believe will hold up in court.
This change in focus and procedure was formalized in published guidelines for conducting compensation analyses. The OFCCP outlined a three-step process that they will be using to analyze compensation data. They have also suggested that contractors adopt this process in their internal analysis.
Creating Similarly-Situated Employee Groupings (SSEGs)
The OFCCP has advised contractors to develop a new grouping method for compensation analysis. These groupings, known as SSEGs, should combine employees with: Similar Work, Level of Responsibility, Skills and Qualifications.
SSEGs should have at least 30 employees, at least five of which should be minorities or females.
Although this definition sounds very similar to the definition of an AAP Job Group, the OFCCP has stated that Job Groups may not be suitable for use as SSEGs. An example of this cited by the OFCCP would be two employees having identical job titles but very different levels of responsibility.
Initial Screening for Potential Discrimination
The initial screening for potential discrimination consists of comparing average salaries of minorities and non-minorities, females and males. A 5% difference in mean salary is a positive indicator that discrimination may exist and shows a potential need to use Multiple Regression Analysis.
Multiple Regression Analysis to Statistically Validate Discrimination
Regression analysis involves identifying all of the independent variables that may affect salary (such as tenure, performance appraisals, education, etc.). A regression analysis should be performed on each of these variables to see if it explains the differences in pay. If disparities in compensation cannot be explained through regression analysis, then the contractor should document its findings and, if necessary, take action to remedy the situation. For many HR and Compensation analysts, multiple regression analysis is a highly complex statistical process that usually necessitates the use of specialized software or outside guidance.
How We Can Help?
HudsonMann will assist your company by providing a compensation analysis that mirrors the procedures used by the federal government. HudsonMann will assist in creating SSEGs, perform the initial screen of your data, and use multiple regression analysis when necessary. Our consultants will help you analyze the results of the analysis so that you are prepared for any OFCCP audits.