Designing a Defensible Municipal Compensation Study

Abstract

Municipal compensation decisions require comparable jobs, consistent pay definitions, and a visible chain from source schedules to recommendations. This methods paper translates WIL-CBM-1.0 into an operational study protocol and reusable collection and review templates. It includes a verified public-source wage example to demonstrate measurement boundaries. It reports no municipal peer-survey results, salary gaps, recruitment effects, or client findings. A public employer can use the protocol to commission a study, evaluate a consultant's deliverables, or prepare its own evidence for review.

The decision to be made

Begin with a written decision question: whether to maintain a salary structure, revise selected relationships, change market positioning, or phase an adjustment within a budget. Define who can authorize the decision and which employees and positions are in scope. A broad question such as “Are we competitive?” is insufficient without the pay concept, comparison market, and effective date.

Keep external competitiveness, internal job relationships, and affordability separate. An employer may be near an external benchmark while retaining compression between supervisory and nonsupervisory roles. Conversely, a documented pay difference does not establish a retention problem. Turnover, applications, vacancy duration, and exit evidence must be analyzed separately before attributing staffing difficulties to compensation.

This paper supplements the published WIL-CBM-1.0 method note [M1]. It does not replace that standard or amend a submitted grant protocol. Its worksheets specify how to document compliance. They are blank instruments, not records of completed engagements.

Study boundary and comparator selection

Freeze the study's job inventory and reference date. Record job identifiers, duties, supervisory scope, credentials, employment status, bargaining unit where relevant, authorized positions, incumbents, vacancies, salary schedules, and scheduled hours. Distinguish a position from a person: an incumbent may hold an assignment that differs from the benchmark description.

Define comparator categories before collecting pay. General municipal operations, utility functions, police, and fire may recruit in different markets and should not be pooled merely because they share an employer. Select peers using service mix, organizational scale, operational complexity, recruitment overlap, governance, and geography. Document why each peer belongs in the category and record both invitations and responses. The client and research team should agree the peer set before collection [M1].

WIL-CBM-1.0 requires at least 10 participating entities per surveyed category [M1]. An invitation is not participation. Maintain a participation register by category, alongside a separate count of usable observations for each matched job. Meeting the category threshold does not ensure that every job has adequate data. If participation falls short, report the gap and continue collection or seek a documented methodological disposition; do not describe the category as compliant or quietly merge unlike categories.

Matching jobs on work performed

Compare duties and responsibility before looking at pay. Prepare a duty crosswalk showing the focal job's functions, the peer job's corresponding functions, supervision, decision authority, credentials, and material working conditions. Preserve the source job descriptions and their dates. Title similarity alone is not a match.

The governing standard permits a reported match at 70% or greater duty overlap; lower-overlap comparisons must be disclosed as below threshold, and blended matches must be explicitly flagged [M1]. Before collection, document how duties will be segmented and how overlap will be assessed. If duties are weighted, state the weights and rationale and retain the unweighted crosswalk. This paper does not prescribe or claim validation of a new numerical scoring instrument.

An independent methods check should review consequential matches, disagreements, and exceptions. A reviewer should be able to reconstruct the overlap decision from the duty evidence. Below-threshold observations belong in the exception register rather than silently entering the primary benchmark. Show sensitivity excluding accepted partial or blended matches when they materially influence the result.

Collecting comparable compensation

Use one observation per peer, benchmark job, pay measure, and effective date. Collect range minimum, midpoint, maximum, and the applicable schedule. Collect incumbent base pay only where lawful and necessary; avoid publishing identifiable employee records. Identify scheduled weekly and annual hours, overtime eligibility, recurring allowances, and employer benefit contributions in separate fields.

Base salary, actual earnings, salary-range midpoint, and total employer cost answer different questions. Compare midpoint with midpoint, or consistently defined incumbent base pay with the same measure. A wage percentile from an occupational survey is not a percentile of municipal salary-range midpoints. Benefits should remain a separate comparison unless valuation assumptions are explicit and comparable across peers.

For police and fire, retain shift cycles, scheduled hours, overtime treatment, and premium pay. Use the actual schedule when annualizing an hourly rate. Do not impose a generic annual-hours assumption on every position. An accounting contribution to a pension plan is an employer expense, not automatically an estimate of the employee's benefit value.

Harmonization and missing data

Retain the original amount, unit, effective date, source locator, and retrieval date. Convert only after resolving scheduled hours and pay frequency. Show any aged amount alongside the original, with the aging factor, source, elapsed period, and formula. Geographic adjustments require a stated source and a reason the adjustment is relevant [M1]. Do not combine aged and unaged values without labeling the comparison.

Use distinct missingness labels for nonresponse, unavailable, suppressed, not applicable, and unresolved. A blank cell must not become zero. Request clarification of suspicious units or outliers and retain the response. Genuine outliers should not be removed merely because they change the benchmark. Document any exclusion before interpreting its effect.

A second analyst should check source-to-row transcription, unit conversions, category participation, and job-match decisions. Report the eligible peer count and usable job-level count alongside every summary. Examine observable differences between participating and nonparticipating peers; an apparently precise median does not eliminate response bias.

Analysis and supporting decision table

Use equal-employer summaries as the default descriptive comparison. If employment weighting is substantively justified, present it separately with the weights and rationale. Do not average different pay concepts. For small matched sets, report the observed range and median with the number and quality of matches rather than imply a representative probability sample.

External pay position = (the employer's specified pay measure / the comparable benchmark) − 1. Express the result as a percentage only when the denominator is positive and the measures are aligned. Compa-ratio = incumbent base pay / the employer's range midpoint. Neither formula independently establishes fairness, legal compliance, or a required adjustment.

Implementation cost is the sum of proposed base-pay changes for included positions, plus separately modeled recurring employer costs. Distinguish filled from vacant positions, annual recurring cost from partial-year cost, and transition expense from ongoing commitments. Record the proposed effective date and assumptions. The compensation decision worksheet accompanies the collection and match templates so that a recommendation can be traced to evidence and cost.

A verified descriptive benchmark: what the number means

BLS QCEW reports $70,262 average annual pay for Florida private construction in 2024 [B1]. The selected cell is NAICS 23, private ownership code 5, annual reference period, all establishment sizes, statewide geography. This is an industry-wide wage measure across occupations. It is not municipal base salary, a construction-manager salary, a salary-range midpoint, or benefit-inclusive compensation.

The accompanying measurement table contrasts this observed benchmark with the uncollected municipal measures. Its purpose is to demonstrate why a correctly sourced statistic can still be the wrong comparator for a particular decision. It supports no municipal pay-gap estimate or salary recommendation. OEWS occupational estimates may supply additional context once occupation, geography, industry group, ownership, vintage, and wage concept are aligned [B3]; this package publishes no OEWS wage finding.

From findings to recommendations

A completed study should give decision makers a job-level evidence table, a match/exclusion register, a missingness summary, and costed options. Each option should state its policy objective, affected jobs, supporting comparison, uncertainty, recurring cost, and implementation implications. The governing body then chooses a policy position; the statistical benchmark does not make that policy choice.

Where evidence is weak, recommend a bounded next step: collect missing peers, resolve duties, separate a mixed job family, or investigate recruitment conditions. Do not fill an evidentiary gap with a market-pay label. Report sensitivity results when the recommendation changes under reasonable match, aging, or benefit assumptions.

Limitations and use

No municipality or peer employer participated in this paper. Municipal duties, schedules, hours, benefits, and staffing outcomes have not been collected. The public benchmark is statewide private-sector context and cannot answer a municipal job-level question. There is no completed peer survey, client experience claim, or empirical validation of a new matching rubric.

The supporting templates are reusable research instruments. Before an applied study begins, the designated methods reviewer and municipal compensation practitioner should confirm the study-specific protocol, collection permissions, disclosure rules, and relationship to WIL-CBM-1.0. Publication authorization remains a separate WIL editorial decision.