Job Evaluation

Two department heads are arguing about whose open role should carry the higher grade, and the only precedents are past arguments. Pay structure built by squeaky wheel eventually surfaces as compression, grievance, or a pay-equity finding — and by then the structure is the defendant.

Milkovich, Newman, and Gerhart's Compensation — the standard text of the field — treats job evaluation as the machinery of internal alignment: define compensable factors (the dimensions of work the organization has decided it pays for), scale each into degrees, weight them, and score every job against the same yardstick. The points produce a worth hierarchy, the hierarchy gets banded into grades, and the grades then meet the market through pricing — a separate step, because job evaluation ranks jobs, it does not price them. Martocchio's Strategic Compensation frames the same machinery as one half of a twin design problem — internal consistency and external competitiveness — and adds the practical stakes: a structure built on stated, consistently applied criteria is also the structure you can defend when compliance questions arrive.

The method's sharpest critic is inside the canon. Lawler's Strategic Pay argues that job evaluation pays the job rather than the person — it can entrench hierarchy, reward job-description inflation, and undervalue the skills the strategy actually needs; his alternative is person-based pay. The critique doesn't retire the method, but it sets its honest terms: point-factor evaluation is systematized judgment, not measurement. The factor plan encodes what the organization chooses to value — choices, not facts — and the method's defense is consistency and transparency of that judgment, never objectivity. Run it for the internal-equity spine, keep the factor plan open to argument, and let market pricing answer the question point totals cannot.

The books describe committees running this over months; here the model judges each job's degree on each factor with a stated rationale, and every calculation — point scales, job scores, the hierarchy, grade banding — runs in code, with compression and bias risks flagged and the market-pricing boundary stated instead of blurred.