Compensating Your Employees Fairly
One book, placed in its field — what Stephanie R. Thomas's book gets right, where it goes further than the rest, and what the rest of the field adds.
Mike West, Editor · Leading With Data · leadingwithdata.guide/guides/profile-compensating_employees_fairly
Arrow keys or swipe to move · N for notes · F for full screen
Compensating Your Employees Fairly
One book, placed in its field — what Stephanie R. Thomas's book gets right, where it goes further than the rest, and what the rest of the field adds.
Mike West, Editor · Leading With Data · leadingwithdata.guide/guides/profile-compensating_employees_fairly
The problem
Detected Pay Disparity erodes when it is left to instinct instead of method.
You were taught the moves piecemeal, never the whole model.
Compensating Your Employees Fairly · Mike West, Editor · Leading With Data
The model
- 01Compensation Policy Clarity and Objectivity
- 02Compensation Data Quality and Completeness
- 03Similarly Situated Grouping Validity
- 04Statistical Analysis Rigor
- 05Detected Pay Disparity
- 06Follow-Up Investigation and Remediation
- 07Perceived Compensation Fairness
- 08Litigation and Regulatory Exposure
Compensating Your Employees Fairly · Mike West, Editor · Leading With Data
Step 1 of 5 · Compensation Policy Clarity and Objectivity
Every pay exception should be traceable to a documented, pre-defined criterion before it is approved.
Misconception
Practitioners believe that giving managers flexibility to reward individuals on judgment produces fairer, more responsive pay than rigid criteria.
Reality
Unstructured discretion is where bias concentrates, not where fairness lives; the same latitude that lets a manager reward a standout also lets unexamined preference set someone's salary. Structure narrows the room where bias operates without eliminating room for merit.
Compensating Your Employees Fairly · Mike West, Editor · Leading With Data
Step 2 of 5 · Compensation Data Quality and Completeness
Total compensation, not base pay, is the correct unit of analysis and must be assembled from multiple systems.
Misconception
Teams assume their HRIS export is analysis-ready and that the numbers in the system reflect what people are actually paid.
Reality
Compensation data is typically scattered across payroll, HRIS, equity, and bonus systems with inconsistent job codes, missing hire-date histories, and stale titles. Assembling a clean cradle-to-grave record is usually the largest and most-skipped part of the work.
Compensating Your Employees Fairly · Mike West, Editor · Leading With Data
Step 3 of 5 · Similarly Situated Grouping Validity
Comparability rests on the substance of the work, not on titles or salary grades.
Misconception
Practitioners think job title or salary grade is a sufficient basis for grouping employees into comparison pools.
Reality
Identical titles can mask genuinely different work, skills, and responsibility levels, while different titles can describe substantively equal jobs. Valid grouping requires evaluating the actual content and demands of the work, not the label attached to it.
Compensating Your Employees Fairly · Mike West, Editor · Leading With Data
Step 4 of 5 · Statistical Analysis Rigor
A pay gap coefficient means nothing without the model specification and assumption checks behind it.
Misconception
That running a regression and reading off the coefficient for gender or race settles whether pay is fair.
Reality
The result depends entirely on which controls you include, whether the model's assumptions hold, and how significance is interpreted; an omitted legitimate variable can create a false gap, and an included tainted variable can hide a real one.
Compensating Your Employees Fairly · Mike West, Editor · Leading With Data
Step 5 of 5 · Detected Pay Disparity
A detected disparity is a signal to investigate, not evidence of wrongdoing.
Misconception
That any statistically significant pay gap is proof of discrimination requiring immediate adjustment.
Reality
A detected disparity is an estimate conditioned on your data, grouping, and model — it flags where to investigate, not a verdict. Statistical significance in a large workforce can mark a trivially small gap, while a meaningful gap in a small group may not reach significance.
Compensating Your Employees Fairly · Mike West, Editor · Leading With Data
A framework to use
Legal Framework for a Disparate Treatment Claim
This framework is part of the full guide. See the guide →
Compensating Your Employees Fairly · Mike West, Editor · Leading With Data
The cheat sheet
- Compensation Policy Clarity and ObjectivityEvery pay exception should be traceable to a documented, pre-defined criterion before it is approved.
- Compensation Data Quality and CompletenessTotal compensation, not base pay, is the correct unit of analysis and must be assembled from multiple systems.
- Similarly Situated Grouping ValidityComparability rests on the substance of the work, not on titles or salary grades.
- Statistical Analysis RigorA pay gap coefficient means nothing without the model specification and assumption checks behind it.
- Detected Pay DisparityA detected disparity is a signal to investigate, not evidence of wrongdoing.
- Follow-Up Investigation and RemediationLegitimate-explanation investigations must be even-handed, not one-sided attempts to explain gaps away.
- Perceived Compensation FairnessFair outcomes and fair-feeling procedures are different levers; you need both.
- Litigation and Regulatory ExposureThe largest litigation risk is a known gap left uncorrected, not the analysis that found it.
Compensating Your Employees Fairly · Mike West, Editor · Leading With Data
Sources
- Compensating Your Employees Fairly — Stephanie R. Thomas
Compensating Your Employees Fairly · Mike West, Editor · Leading With Data