OFCCP Compliance Tip: You Should Be Doing Your Statistical Disparity Analyses By “Sub-Groups” and Not By “All-Minorities” vs “All Non-Minorities”
In addition to our Week In Review (WIR) and OFCCP Compliance Alerts, we periodically share Compliance Tips. These posts are snippets of advice for federal contractors, provided by a variety of field experts. Today’s post comes to us from John C. Fox, Esq., President and Senior Partner at Fox, Wang & Morgan P.C.
Today’s OFCCP Compliance Tip: You Should Be Doing Your Statistical Disparity Analyses By “Sub-Groups” and Not By “All-Minorities” vs “All Non-Minorities”
To start the New Year off right, let’s fix a basic error many HR consultants and law firms routinely make as to one of the most important statistical analyses you must annually undertake as part of your AAP for Minorities and Women: “Disparity Analyses” for Hires, Promotions and Involuntary Terminations.
After 15 years of doing it incorrectly, OFCCP last year (finally) changed the (erroneous) approach the agency first adopted in 2000. We also understand OFCCP has now delivered new analytics software to its Compliance Officers allowing for the proper “sub-group” analyses. Two adverse decisions OFCCP suffered in the same week in both the Cargill and Jeanswear cases in 2014 drove OFCCP’s change of direction and caused OFCCP to (finally) comply with basic Title VII law which the agency had stubbornly avoided since it began requiring Disparity Analyses in 2000. In both of those case decisions, U.S. Department of Labor Administrative Law Judges held that OFCCP’s approach to aggregating all “Minorities” together (i.e. Blacks, Hispanics, Asians, Native Americans and Hawaiians) and comparing their statistical percentage of rejection (let’s say for those considered for hire: typically known as an “Applicants vs Hires” comparison) against the rejection percentage for all Whites, was in error and did not properly interpret and implement Title VII law.
NOTE: Technically, one compares “rejection” percentages and not “hiring percentages” (since you are measuring and comparing the frequency of “adverse action.” However, the percentages of hires and rejections are reciprocal, so many (improperly) compare hiring percentages and get to the same conclusion as if they had done it correctly by analyzing rejection for hires percentages.
FIRST: The Courts in both Cargill and Jeanswear adopted traditional Title VII law principles calling for contractors and OFCCP to first identify the “Most Favored Group” (“MFG”) from among the six groups federal law presumptively protects—meaning the protected group from among all those the law protects which enjoyed the lowest percentage of rejection or “adverse action” (i.e. the group with the lowest percentage of rejections for Hire (or viewed backwards from a “selections” point of view, the group with the highest percentage of hires), or the lowest percentage of rejections for Promotions (in the case of a promotions analysis), or those fired less frequently on a percentage basis (now you see why we technically should analyze adverse action and not selections…those with the greatest percentage of selections for termination is not the “Most Favored” Group!!!).
SECOND: The contractor/OFCCP must then compare the percentages of rejection of the MFG against the percentages of rejection of EACH of the other five protected groups (popularly and informally called “subgroups”) which the law protects. (The six “subgroups” federal law currently presumptively recognizes and protects are: Whites, Hispanics, Blacks, Asians, Native Americans and Hawaiians). Of course, OFCCP could undertake national origin discrimination analyses in its discretion, even though contractors are not required to do so. Example: American manager Applicant rejections verses Japanese manager Applicant rejections at a stateside establishment of a company headquartered in Japan.
Example: If Hispanics were the MFG, one’s hiring Disparity Analyses would look like this:
- Percentage of Hispanic (Applicants) rejected vs. the percentage of White (Applicants) rejected
- Percentage of Hispanic (Applicants) rejected vs. the percentage of Black (Applicants) rejected
- Percentage of Hispanic (Applicants) rejected vs. the percentage of Asian (Applicants) rejected
- Percentage of Hispanic (Applicants) rejected vs. the percentage of Native American (Applicants) rejected
- Percentage of Hispanic (Applicants) rejected vs. the percentage of Hawaiian (Applicants) rejected
Your Disparity Analyses for Hires would NOT look like this anymore (they never should have, but most of you uncritically followed OFCCP right off the cliff’s edge Pied Piper-style):
- Percentage of Minority (Applicants) rejected vs. the percentage of White (Applicants) rejected; or
- Percentage of Hispanic (Applicants) rejected vs. the percentage of all others (Applicants) rejected (i.e. Whites+ Blacks + Asian + Native American + Hawaiian) (what OFCCP used to call a “Non-Hispanics” analysis, or a “Non-Black” analysis or “Non-Asian” analysis…meaning “Non-Hispanics” or “Non-Blacks” or “Non-Asians” were the “victims” in before the Cargill and Jeanswear decisions.
If you are still doing “All Minority” analyses, or your consultants are doing so on your behalf, JUST STOP! Don’t do them. Irrelevant. Data clutter. Waste of time. Tells you nothing of legal significance. You are unnecessarily wasting precious resources. STOP IT!
HAPPY NEW YEAR! You are saving money and resources already! Be careful out there!
-John C. Fox