Review of the Proposed OFCCP Standards for Assessing Systematic Compensation Discrimination

As noted in a previous posting, the OFCCP is now proposing new standards for assessing Systematic Compensation Discrimination. While there are some interesting deviations from the OFCCP’s previous practices in this regard, what is most interesting is that the agency is now proposing a standardized method for this determination. I review the basic points made in this document, and offer some suggestions for employers, in light of these proposed standards.

Asserting that the OFCCP had not issued previous standards for assessing systematic compensation discrimination, the agency states that an informal, controversial approach known as the “pay grade theory” was often used. As explained in these proposed guidelines,

the “pay grade theory” assumed that employees were similarly situated and hence comparable, if the company had placed their jobs in the same pay grade. In one version of the “pay grade theory,” the median pay of males versus females, and minorities versus non-minorities could then be compared. Significant differences would be examined, and a determination made as to whether significant differences could be “explained” by median differences in other factors, such as education, experience, or performance. Inability to explain a discrepancy would be viewed as discrimination. A somewhat different version of the pay grade theory used pay grade as one factor in a multiple regression analysis model that typically was applied to all exempt employees in a single, pooled analysis. This statistical model was then used to determine whether there was discrimination.

The proposed guidelines, simply put, argue that the mere fact that several employees are in the same job grade is not sufficient reason to assume they are comparable and that they are similarly situated. As explained in these guidelines, the “pay grade theory” fits with the comparable worth theory, which has been rejected by the courts. The argument made in these proposed guidelines is more similar to the Equal Pay Act of 1963. Specifically, the proposed guidelines would define employees as “similarly situated” when they are performing similar work, have similar responsibility levels, and occupy positions involving similar qualifications and skills.” Accordingly, OFCCP would investigate whether employees are similarly situated by looking at job descriptions and conducting employee interviews.

Once it is determined which employees are similarly situated, these proposed guidelines state that the “legitimate factors” that affect pay would be determined, such as education, prior work experience, and performance ratings. These guidelines state that OFCCP will investigate the facts to ensure that these factors are “legitimate” and are not influenced by unlawful discrimination. But the judgment of whether a factor is influence by unlawful discrimination will be based on the particular facts of the case. Once the relevant factors are determined, the proposed guidelines state that multiple regression analysis will be used to assess discrimination. Standard statistical indices (e.g., statistical significance tests) will be used towards this end.

Despite the heavy reliance on statistical tests, the proposed guidelines appear adamant regarding the need for anecdotal evidence as well. As stated in these proposed guidelines, “Except in unusual cases, OFCCP will not issue a Notice of Violation (NOV) alleging systematic compensation discrimination without providing anecdotal evidence to support OFCCP’s statistical analysis.” The reverse is also true; the proposed guidelines state that only in unusual cases will the OFCCP assert that there is systematic pay discrimination without statistical evidence.

Finally, it is interesting to note that these proposed guidelines indicate that OFCCP could assert pay discrimination is occurring if the contractor establishes pay rates for jobs that are occupied primarily by women or minorities that are “significantly” lower than rates established for jobs that are primarily occupied by men or non-minorities, where there is evidence suggesting intent. The proposed guidelines continue by stating that intent in this situation may be demonstrated by showing that a market survey was used, but that the company assumed a below average position for the jobs primarily occupied by women or minorities, and a “full market rate” for jobs primarily occupied by men or non-minorities, under the same market survey. This position suggests that greater care needs to be taken in how salary surveys are used and for covered organizations to be more cautious in how pay rates are established.

The good news appears to be that more standardized procedures for assessing discrimination are around the corner, which I believe, will make it easier for covered companies to defend their pay practices in an OFCCP investigation. Comments on these proposed guidelines may be sent by December 16, 2004 to the OFCCP.

Some suggestions for covered companies:

1. Covered companies will need to be more proactive in using multiple regression analysis in assessing their compensation data;
2. Make sure that pay discrepancies can be readily explained by legitimate factors. Be sure that the legitimate factors cannot be swept aside by OFCCP as being discriminatory in themselves;
3. Covered companies may need to do a better job of communicating to employees how pay decisions are made
4. Have a process for identifying possible complaints about pay (e.g., use employee satisfaction surveys), rather than letting them fester on;
5. Have an internal grievance process to address possible complaints about pay.