Anonymize equal-pay audit data for compensation-equity analysis – CCPA/HIPAA-compliant de-identification per Title VII §2000e-2
Pay-equity audits conducted to identify potential violations of Title VII §2000e-2 require linking individual employee compensation figures to protected-class characteristics. anonym.legal pseudonymizes employee identifiers in compensation datasets so external equity analysts and outside counsel can perform regression modeling and pay-disparity analysis without unnecessary access to individually identified salary records.
When this applies
Apply this workflow before sharing compensation datasets with pay-equity consultants, econometricians conducting regression analysis, or outside counsel evaluating equal-pay exposure under Title VII.
How anonym.legal handles it
- Upload compensation data exports from your HR or payroll system to anonym.legal.
- The engine identifies and pseudonymizes employee names, employee IDs, and manager identifiers in the dataset.
- Protected-class fields — gender, race, ethnicity — are retained in their original coded form for regression analysis.
- Compensation fields — base pay, bonus, total remuneration — and job-level fields are retained for equity modeling.
- The pseudonymized dataset is exported in the original format for consultant or counsel use.
- A reversible mapping key is stored for re-identification of individuals identified as potential pay-disparity outliers requiring remediation.
What you provide
- Compensation data exports from HR or payroll systems in CSV or XLSX format
- Job-level and job-family classification structure
- Protected-class field definitions and coding scheme
Limitations & cautions
- anonym.legal does not perform pay-equity regression analysis; statistical modeling must be conducted by a qualified compensation analyst or economist.
- Retaining protected-class fields alongside compensation figures carries inherent re-identification risk that pseudonymization of names alone cannot fully eliminate in small demographic cells.
- The Equal Pay Act (29 USC §206(d)) and state equal-pay statutes may impose overlapping but distinct obligations; this workflow is framed around Title VII §2000e-2 and does not constitute EPA-specific guidance.
- Pay-equity analysis results may be attorney-client privileged or attorney-work-product; privilege designation must be managed separately from this pseudonymization workflow.
FAQ
Can this workflow accommodate multiple job-family structures across different business units?
Yes. The pseudonymization applies to all rows in the dataset regardless of job family or business unit. Job-family and business-unit codes are treated as structural fields and are retained for cross-unit equity analysis.
Will the pseudonymization affect the statistical validity of a pay-equity regression model?
No. Removing names and employee IDs does not affect the numeric compensation values or the protected-class and job-level variables that drive regression results. The pseudonymized dataset is statistically equivalent to the original for modeling purposes.
Is this workflow suitable for preparing data for an OFCCP compliance evaluation?
Pseudonymization is an internal preparation tool; actual OFCCP submissions must include identified employee records as required by the agency. Use this workflow for internal pre-submission analysis and attorney review, not for the submission itself.