Turnover Rate vs. Absence Rate: Definitions, Denominators, and Limits

The calculations answer different descriptive questions

HR Hub's selected-separation rate adds documented voluntary and involuntary departures for a stated period and divides by a two-point average of starting and ending headcount. The result describes included departure events relative to that workforce denominator; it does not follow one starting cohort and is not the same as retention.

The equal-schedule absence scenario divides selected absence days by entered employees multiplied by available workdays per employee. It describes selected time lost under one modeled schedule. When schedules differ, a validated aggregate scheduled-days or FTE-days denominator is required outside the tool.

Comparisons require separate counting policies

Turnover requires a population, date range, separation-event policy, category definitions, and headcount method. Absence requires a population, date range, unit, schedule denominator, and rules for approved, protected, disability-related, occupational, or other leave categories.

The rates cannot be compared as if a percentage point has the same meaning in both measures. Trend each metric against its own prior, scope-aligned periods and preserve the source records used for numerator and denominator.

Neither rate diagnoses cause or determines action

A change in turnover does not prove a compensation, management, labor-market, or employee-experience cause. A change in absence rate does not prove attendance misconduct, health status, disengagement, inadequate staffing, or lost service coverage.

Use the rates to identify where records need review. Investigate role, location, tenure, schedule, leave classification, exit evidence, workload, and policy context separately, with appropriate privacy, accommodation, leave, labor, and employee-relations review.

Method, scope, and sources

Both views use a publisher-defined selected-separation rate: included voluntary plus involuntary departures divided by two-point average headcount for the stated period. The advanced workspace adds a comparable prior-period calculation, category shares, a simple monthly/quarterly annualized run-rate, and a separate replacement-cost sensitivity based on a user-entered amount per included departure.

Define the population, dates, separation categories, and headcount method before use. The annualized value is an arithmetic projection, not an observed annual rate. The cost sensitivity is not part of the turnover rate and is not audited cost or savings. Retirements, transfers, deaths, disability exits, rehires, contractors, and internal moves require an explicit classification policy.

Method status: Publisher-defined.

Worked example

Hypothetical quarterly scenario: a defined unit has 100 starting and 104 ending employees, with three resignations and one employer-initiated termination under its documented policy. Selected-separation rate = 4 ÷ 102 × 100 = 3.92%. A simple annualized run-rate would be 15.69%, but it is only a constant-rate projection. Any per-departure cost remains a separate organization-specific sensitivity.

Sources

• U.S. Bureau of Labor Statistics — Job Openings and Labor Turnover Technical Note: Directly supports separations ÷ employment × 100 and the distinct quits, layoffs/discharges, and other-separations taxonomy. It defines quits as voluntary except retirements/transfers, and identifies layoffs/discharges as employer-initiated.
• Global Reporting Initiative / Global Sustainability Standards Board — Standards interpretations: Confirms GRI 401 reporters are not required to use end-of-period employee numbers to calculate turnover rates from 1 July 2018; it does not prescribe the calculator's two-point denominator.

Method, scope, and sources

Equal-schedule days scenario: selected absence days ÷ (available workdays per entered employee × entered employee count) × 100. The advanced view also shows selected days per entered employee, an arithmetic available-days residual, and a matched-reference delta; these are descriptive publisher-defined outputs.

Define the population, period, unit, denominator method, and included approved, protected, or other absence categories. Use aggregate scheduled or FTE days when schedules differ. The residual does not measure attendance, work performed, replacement coverage, or service coverage. Compare a reference rate only when population, period, unit, numerator, denominator, and inclusion rules match. The result does not establish cause, misconduct, performance, or action.

Method status: Partially supported.

Worked example

Hypothetical equal-schedule scenario: 85 entered employees each have 22 available workdays in the same month, giving 1,870 modeled available days. Forty-two selected absence days produce 2.25%. Selected days per entered employee = 0.49, and the arithmetic residual is 1,828 days; that residual is not proof of attendance or coverage.

Sources

• Office for National Statistics — Sickness absence in the UK labour market: 2025: Defines sickness absence rate as working hours lost because of sickness or injury divided by total hours × 100, and distinguishes it from days lost per worker.
• NHS England / NHS Digital — NHS Sickness Absence Rates, February 2024 — Methodology: States that sickness absence rates divide FTE number of days sick, including non-working days, by FTE number of days available; it is the closest direct days-based support but is program-specific.
• Statistics Canada — Work Absence Rates: Data quality, concepts and methodology: Documents an hours-lost personal-reason measure with stated exclusions and distinguishes incidence from time lost, supporting explicit scope and comparability rules.