For purposes of all analyses that are done in conjunction with updating an AAP, and/or submitting data to the OFCCP in the course of a compliance review or complaint investigation, it is very important to associate ALL applicants with each selection decision for which they were considered.* In your example, the hiring decision was made in 2020 so all applicants -- no matter when they applied -- for that opportunity must be included in the analysis of that opportunity. It may help (in a way!) to think of each final selection decision (whether for hire or competitive promotion/transfer) as resulting in a discrimination complaint. When a selection decision is challenged (or could be!) with an allegation of discrimination the employer will definitely want to include every person it considered for that decision -- all of them competed WITH EACH OTHER -- AND ONLY WITH EACH OTHER -- for that PARTICULAR opportunity. For good or ill...no fudging the numbers! Knowledge is power and the employer NEEDS to know -- well before the OFCCP is told or finds out! whether its selection criteria are being applied differently on the basis of race or gender (disparate treatment) or whether its selection criteria even uniformly applied are excluding women at faster rate than men, for example (disparate impact). Doesn't it make sense that an employer can only make a decision FROM AMONG THE CHOICES IT HAS AT THE TIME OF THE DECISION? So, persons who are applicants in 2020 even if for the same job title are not the people whose qualifications should be "compared" with the qualifications of the persons who were selected/rejected in 2019 from among people in those applicant pools. You will of course have recognized that ANY aggregation over time will combine people who didn't compete WITH EACH OTHER. IF statistical analyses show [statistically] significant differences in selection rates the very best the employer can do is to ensure that the data will permit it to make a "deeper dive" into applicant pools for each selection decision in order to investigate and – hopefully – defend the decision(s) on the basis of job related criteria. Employers DO NOT want to be in the position of trying to persuade the OFCCP that the analyses they submitted/the OFCCP investigated are based on the wrong data and can we pretty please have a "do over".** Aggregation of selection decisions should occur ONLY for opportunities having THE SAME SELECTION CRITERIA, irrespective of job title, much less job group or any other unit. Aggregation, even at the same job title level, will always be an imperfect basis for analyzing disparate TREATMENT and will, very often, also be imperfect for analyzing disparate IMPACT. I'm willing to bet that your professional experience is that selection criteria are not always identical depending on, for example, how anxious the employer is to fill a vacancy, or whether the person actually making the hiring/promotion decision is strict in adhering to qualification requirements -- or not so strict. But the very least the employer can expect the OFCCP to accept is that applicant pools be matched to the AAP Year IN WHICH THE SELECTION DECISIONS BEING ANALYZED OCCURRED. And the least the contractor should be able to expect from the agency itself is that it also will not diddle the numbers (e.g. by aggregating across job titles or across AAP years where more than one year of data have been submitted) in order to achieve the result IT is trying to find. You can't manage what you don't know so strive to really KNOW the impact of your selection criteria and to KNOW that if any race/ethnic/gender group is being selected at a lower rate than ANY OTHER GROUP and to KNOW WHICH selection criterion(a) is causing the impact AND to KNOW that there is a demonstrable "business necessity" for such selection criterion. How can the employer possibly KNOW unless it is analyzing the data that include all -- and only -- the persons it "considered" for a PARTICULAR job. It's one thing to know that the requirement of a degree in engineering is excluding Hispanics at a faster rate than Whites, for example, than to FIND OUT TOO LATE that the requirement of a college degree (in anything) to be promoted to exempt level -- or even management -- is excluding Blacks, Hispanics, Native Americans and Whites at a faster rate than Asians. Think about how you will demonstrate that a Bachelor's degree in any subject matter -- is "job related" to ANY job. ESB Note: * Although space does not permit a full discussion here, please be aware that these same principles apply when analyzing terminations that are a result of a reduction in force. These must be analyzed using ONLY the pools of people WHO WERE AT RISK for termination. That will include reductions made by department and/or by job title. (I’ve never met an employer who RIF’d by AAP Job Group!!!) Who were the people who competed with each other TO KEEP THEIR JOBS? The persons not selected for retention are the equivalent of “rejected applicants” in the more typical analysis of applicants versus hires/competitive promotions. They are the denominator in your analysis; the people who survived the RIF are the numerator. **As embarrassing and difficult as it is to try to disavow one’s own data/analyses/admissions (!) the employer who finds itself in this situation should seek experienced legal counsel and/or consultants to assist in revising its data and in persuading the OFCCP that it should accept these revisions. In my view it doesn't help that all too many systems are geared to -- even CALLED -- "Applicant Tracking Systems". Staffing managers (to whom these systems typically belong) have a legitimate need to "track" applicants in order to manage -- and evaluate -- the productivity of their department. Compliance managers (and CFOs and stockholders!) need to be focused on applicants ONLY as they are related to DENIAL of the opportunities WHICH THEY SOUGHT. Because that's what the OFCCP is interested in! Think about whether your automated system helps or hinders you in analyzing/monitoring employer Selection Decisions. You might also like to read some of the articles in the OFCCP Digest about "Counting the Right Numbers", the "law of big numbers" and other topics having to do with discrimination analysis. How the employer properly retains, extracts and aggregates data it reports to the OFCCP -- or improperly does these things -- has enormous potential consequences for the bottom line. Best wishes! Ellen Shong Bergman

You can use this OFCCP audit checklist to ensure you're doing what is required to maintain OFCCP's regulations including VEVRAA, Section 503, and EO 11246. Or request a demo to streamline your compliance and recruiting efforts.