Applicant flow data is where non-discrimination theory meets hiring operations. Every company knows it should hire fairly; applicant flow data is how companies prove it, or, more uncomfortably, how they find out they aren't. The data tracks every candidate through every stage (applied, screened, interviewed, offered, hired) and breaks the numbers down by demographic categories. When the pass-through rates at any stage diverge sharply between groups, that's a signal to investigate. The Four-Fifths Rule, which the EEOC has used since the 1970s, gives a rough threshold: if a protected group's selection rate falls below 80% of the highest group's rate, the hiring practice may be considered to have disparate impact.
What Applicant Flow Data Includes A full applicant flow dataset captures several dimensions per candidate. Position applied for (the specific requisition). Application date. Self-identified demographics (race, gender, disability status, veteran status, collected voluntarily). Stage progression through each step of the hiring process. Final disposition (hired, rejected, withdrew). Reason for disposition where available.
Demographic data is collected through the voluntary self-identification process on applications. Per EEOC guidance, the data collection must be voluntary, separated from the selection decision, and used only for aggregate analysis and reporting.
How Applicant Flow Data Supports Disparate Impact Analysis The core analysis looks at selection rates at each stage. If 1,000 candidates apply, 200 are White, 150 are Black, and the remainder are other races, and the employer hires 40 White and 20 Black candidates, the selection rates are 20% and 13.3% respectively. Black candidates' selection rate is 66.6% of the White rate, which falls below the 80% threshold under the Four-Fifths Rule.
That result doesn't automatically establish discrimination . Employers can rebut the finding by showing the practice is job-related and consistent with business necessity. But the data triggers the conversation and creates obligation to examine the practice.
Is Demographic Self-Identification Mandatory? For applicants, no. Self-identification is always voluntary. For federal contractors, the invitation to self-identify is required (both pre-offer for veterans and disability, post-offer for race and gender), but the applicant's response is voluntary. Declining to answer is a valid response and doesn't penalize the candidate.
How Employers Actually Collect Applicant Flow Data Most employers use an applicant tracking system (ATS) that includes voluntary self-identification during the application flow. The data lives in the ATS and gets exported for quarterly or annual analysis. The invitation language, the timing of the request, and the way the data is separated from selection decisions are all governed by OFCCP and EEOC guidance.
Common collection mistakes include combining identification data with the application form itself (which creates the appearance that demographics influence decisions), failing to track withdrawals (which skews the denominator), and applying different data collection standards across business units (which makes cross-unit analysis meaningless).
Using Applicant Flow Data to Actually Improve Hiring Data without action is decoration. Four practices convert applicant flow data into better outcomes. First, run the analysis quarterly rather than waiting for audits. Second, investigate every disparate pass-through rate with a named owner and a specific hypothesis. Third, examine the qualification filters, interviewer panels, and structured assessments to find where the disparity actually enters the process. Fourth, close the loop by documenting the changes made and measuring the subsequent data.
Strong applicant flow programs also tie back to broader onboarding and employee retention data. Hiring fairly is the start; keeping a diverse workforce employed and engaged is the longer game. OFCCP expects to see disparity analysis, documented investigations, and corrective actions in federal contractor audits, and the same discipline serves non-contractor employers just as well even when the legal obligation is lighter.
.svg)
.svg)
