← The library

Absence management (contested) Contested practice

Bradford Factor

The Bradford Factor scores absence as spells squared, times total days, so frequent short absences score far higher than one long one.

It gives absence management a number that looks objective, and that is exactly the danger: the formula cannot tell a sick child from a sick culture, or a disability from a discipline problem.

Problem
Absence management (contested)
Altitude
Function
Effort to run
Light
Evidence base
Emerging

Theory & origin

Named after the University of Bradford's School of Management, the factor is built on an operational observation: five separate Mondays off disrupt a team far more than one five-day flu does, so spells get squared while duration only counts linearly. HR systems adopted it because it turns messy absence patterns into a single, sortable number with policy trigger bands. The controversy is structural, not cosmetic. The formula is blind to cause. Chronic illness, disability-related absence, and caring responsibilities all produce high-frequency patterns, which is how a mechanical Bradford trigger walks an employer straight into a disability-discrimination claim, and walks employees into hiding illness instead, which costs more than the absence ever did.

Key components

The parts of the model and what each one means, in plain terms.

Spells (S)
Separate occasions of absence, regardless of length. This is the squared term. The formula's entire opinion is that frequency disrupts a team more than duration does.
Days (D)
Total days absent across those spells. This term only counts linearly.
Score = S² × D
Five one-day spells: 5 squared times 5 equals 125. One five-day spell: 1 squared times 5 equals 5. Same number of days, 25 times the score.
Trigger bands
Policy thresholds for an informal chat, a formal review, or a capability process. This is where the formula has to stop and human judgment has to start.

Explore the model

2² × 4 = 16

spells squared, times days

Spells (S), separate occasions2
Days (D), total days absent4

How a consultant runs it

  1. 01 Use the score as a trigger for a conversation, never as an automatic sanction. The formula flags a pattern. Humans establish the cause.
  2. 02 Exclude protected and declared absences, like disability-related leave, pregnancy, and caring leave, from the calculation before any band gets applied.
  3. 03 Set trigger bands with employee representatives and publish them. A secret formula breeds exactly the distrust it is supposed to measure.
  4. 04 Read high scores as diagnostics first. Clusters under one manager or one shift pattern are a management finding, not twelve separate individual ones.
  5. 05 Track presenteeism alongside it. If scores fall while sick colleagues are visibly working through illness, the policy is succeeding at exactly the wrong thing.

When to use

  1. 01 Surfacing absence patterns a manager's memory alone would miss, across a large workforce
  2. 02 Shift and frontline operations, where short-notice absence has real coverage costs
  3. 03 As one input into a supportive attendance conversation, with the cause established before any consequence

When not to use

  1. 01 As an automatic disciplinary trigger. Mechanical application is the most direct route to a discrimination claim.
  2. 02 On populations with declared disabilities or caring duties, unless you build in exclusions. The formula punishes the exact pattern their situation produces.
  3. 03 Where the real problem is a sick culture. High scores clustered under one manager are not twelve separate attendance problems.

Worked example

A 900-person contact center runs Bradford bands mechanically and opens capability cases on everyone scoring over 500.

The audit finds one case is migraine-driven, a likely disability accommodation failure that gets quietly settled, and that 60% of the high scores sit in two teams with the same shift pattern and the same team leader. The formula was punishing individuals for what was really a rota problem.

The rebuilt policy excludes protected absences, uses scores to trigger supportive conversations instead of sanctions, and redesigns the rota. Absence falls 22% the next year, almost all of it from the two teams whose real problem was never the employees.

Common pitfalls

  1. 01 Firing automatic sanctions off the raw score. This is the pattern behind most Bradford-related tribunal losses.
  2. 02 Counting disability, pregnancy or caring absences because nobody built the exclusions in
  3. 03 Driving presenteeism, where employees work sick to protect their score, infecting the team and costing more overall
  4. 04 Reading individual scores when the real signal is a managerial or structural cluster

Sample deliverable

One real engagement, start to finish. Watch the numbers travel from raw input, onto the chart, into the finished artifact.

Absence pattern scan: Jakarta contact center

Input

  • Case A: 8 spells × 10dscore 640
  • Case B: 4 spells × 6dscore 96
  • Case C: 2 spells × 8dscore 32
  • Case D: 1 spell × 20dscore 20
↓

Process

Spells and days convert into a score, and scores are read against bands, then against causes

↓
OutputDeliverable

Absence pattern scan: Jakarta contact center

  • Case Amigraine, needs accommodation not discipline
  • Cluster60% of high scores, one rota
  • Fixexclusions plus rota redesign, absence down 22%

Sources

Next in the library Performance Improvement Plan (PIP) →