OFCCP’s Proposal To Rescind The 2006 Systemic Compensation Analysis Standards: What Are They Thinking? – By Dr. Jim Higgins, Ed.D.

On January 3, 2011, the Office of Federal Contract Compliance Programs (OFCCP) announced (see 41 CRF Parts 60-1 and 60-2) that it is proposing to rescind two guidance documents addressing the methodologies to be used by federal Contractors to identify and rectify compensation discrimination. These two documents, established under the George W. Bush Administration, include the Interpretive Standards for Systemic Compensation Discrimination and the Voluntary Guidelines for Self-Evaluation of Compensation Practices for Compliance.

According to the news release, their stated reason for proposing to rescind the compensation analysis guidelines and the methodology that has been utilized since 2006 is that they “limited OFCCP's ability to effectively investigate, analyze and identify compensation discrimination.” This is an extraordinary claim, especially since the release fails to address the primary question. What exactly does it mean to “effectively” investigate, analyze, and identify compensation discrimination?

How this question is answered is critical. If an “effective” analysis implies finding pay disparities that must be corrected, it then implies that any disparity is more likely than not caused by discrimination. Therefore any methodology that fails to find many examples of discriminatory pay practices on the part of federal Contractors is undesirable to the OFCCP. From this perspective, the current compensation analysis strategy is not effective because the majority of employee differences appear to be explained by legitimate differences in employee characteristics rather than contractor bias in setting pay.

If, however, an “effective” analysis means that only those instances where pay disparities are not explained by legitimate differences in job related employee characteristics, it is difficult to think of an analytical strategy that would be more effective than that which was established in 2006. Multiple regression, as a statistical test is unique in the fact it is perfectly designed to identify which variables, having controlled for other variables, influence a criterion (e.g., employee pay). Multiple regression identifies how much a variable like gender or minority status impact pay “after” controlling for legitimate job-related factors. Multiple regression allows you to calculate how much an employee “should” be paid if the same objective factors were applied to all employees equally. Multiple regression allows you to even determine how much of employee compensation is accounted for by your explanatory variables. Finally, multiple regression allows you to find whether group differences exist and then to identify specifically how much each impacted employee should be paid in order to remove statistically significant differences in pay.

Since I find it difficult to believe the OFCCP wants to indentify discrimination where none exists, I can only presume that there must be some other reason for their desire to rescind the 2006 guidelines. It can’t be that the statistical test is wrong. If that were true and multiple regression analysis was an ineffective tool for modeling pay practices, it would imply that weather forecasters are using the incorrect tool to predict the weather, companies are using the wrong statistics to model fuel and other commodity prices that guide their hedging/purchasing decisions, medical doctors are using the incorrect tool to allocate livers for liver transplantation, and personnel psychologists use the wrong tool to create composite scores to predict employee performance.

One of the hallmarks of the set of guidelines as created in 2006 is the care with which they were designed. Steps were taken to choose the correct statistical test, encourage contractors to use it correctly, and create a multi-step process designed to actually “understand” observed group differences in pay.

What is the alternative? If one picks and chooses the occasional employee who may earn less or more than his or her peers and relies solely on anecdotal information, the likely result will be contractors chasing up and down a multitude of blind alleys until they get so tired to answering questions that they simply give up and do whatever they are told so that they can get on with their core business. I certainly hope this is not the route that will be followed as it will place the valuable and lofty goal of affirmative action and equal employment opportunity in a position where people will simply comply with little desire for honest results.

There is nothing about the 2006 compensation analysis guidelines that limits the OFCCP’s ability to investigate, analyze and identify compensation discrimination. The guidelines do not preclude the OFCCP using other analytical techniques. It is possible, therefore, that what the OFCCP is really trying to address is the discovery that the technical sophistication required to use multiple regression as an enforcement tool is beyond the skill level of most OFCCP Compliance Officers. Further, it is possible that once large numbers of employers had conducted valid analyses of their compensation practices it turned out that in the majority of cases differences in pay between males and females and minorities and non-minorities are actually due to legitimate factors and not discrimination. Because this calls into questions many of our previous assumptions, enforcement agencies may have decided that, contrary to the evidence, the analytical methodology developed in 2006 “must” not be right.

There are a lot of assumptions that are made when we talk about how federal Contractors approach compensation discrimination. Many assume that contractors actually want the guidelines to be rescinded. Many others assume that contractors have not even implemented the guidelines to start with. Still others assume that contractors do not want the OFCCP to provide guidance with respect to the best way to conduct compensation analysis.

As a way to help provide information from the contractor community to the OFCCP, Affirmative Action Services (www.AffirmativeActionServices.com) is conducting a survey of federal Contractors to collect information about how they perceive the current guidelines which are based around multiple regression as well as how effective they find them, how burdensome they believe them to be, and how they would like to see them changed. In late February, Affirmative Action Services will analyze the data collected from this survey and conduct a webinar with those who completed it. Afterwards, they will provide the analysis to the OFCCP in a non-partisan and objective scientific manner for the OFCCP to use as they will as they consider what changes to make to how they analyze contractor pay practices.

If you represent a federal contractor, you are invited to complete this important survey. If may be found at the following link: http://www.surveymonkey.com/s/OFCCP_Survey.

If you wish to provide feedback directly to the OFCCP, your comments are due by March 4, 2011. Comments should be submitted, identified by the number 1250-ZNE, according to any of the following instructions:

1) Log in to the Federal eRulemaking Portal: http://www.regulations.gov and follow the instructions for submitting comments.
2) Fax: (202) 693-1304 (for comments of 6 pages or fewer).
3) Mail: Director, Division of Policy, Planning, and Program Development
Office of Federal Contract Compliance Programs, Room N3422
200 Constitution Avenue, NW.
Washington, DC 20210.

Note that receipt of submissions will not be acknowledged. However the you
may request confirmation that a submission has been received by telephoning the OFCCP at (202) 693-0102 (voice) or (202) 693-1337 (TTY). Note that these are not toll-free numbers.

All comments received by the OFCCP, including any personal information provided, will be available online at http://www.regulations.gov and for public inspection during normal business hours at Room C3325, 200 Constitution Avenue, NW., Washington, DC 20210. Individuals needing assistance to
review comments will be provided with appropriate aids such as readers or print magnifiers.

FOR FURTHER INFORMATION CONTACT:
Director, Division of Policy, Planning, and Program Development
Office of Federal Contract Compliance Programs
200 Constitution Avenue, NW., Room N3422,
Washington, DC 20210.
Telephone: (202) 693-0102 (voice) or (202) 693-1337 (TTY).

If you have any questions about how to analyze your compensation practices, contact Dr. Jim Higgins at JHiggins@AffirmativeActionServices.com or visit our website at www.AffirmativeActionServices.com.