Leave & Absence
What is the Bradford Factor and how is it calculated?
Last reviewed 4 May 2026
What it is, where it came from
The Bradford Factor was developed at the Bradford University Management Centre in the 1980s as a way to quantify the disruption caused by an employee's pattern of sickness absence — not just the total days lost, but the shape of those days. The premise: frequent short absences are more disruptive than occasional long ones.
Forty years later, the Bradford Factor is the most widely used absence-scoring tool in UK HR. It's also one of the most criticised — for sound reasons we'll come to.
The formula
Bradford Score = S² × D
where S = the number of separate spells of absence in the period, and D = the total days absent.
The squaring of S is the key feature: it weights frequency disproportionately compared to duration.
Worked examples
Same total days, very different scores:
| Pattern | S | D | Score (S² × D) | |---------|---|---|----------------| | One 10-day absence | 1 | 10 | 10 | | Two 5-day absences | 2 | 10 | 40 | | Five 2-day absences | 5 | 10 | 250 | | Ten 1-day absences | 10 | 10 | 1,000 |
A worker with ten one-day absences scores 100× higher than one with a single 10-day absence — even though both have lost the same total days to the employer.
The penalty for frequency, not duration, is what makes the Bradford Factor controversial: it rewards staying off longer rather than coming back too soon.
The reference period
Most policies use a rolling 12-month period. Each new absence adds to the count; absences that fall out of the 12-month window stop counting.
Some employers use a fixed leave year instead — easier to administer but creates a "race" effect where workers near year-end may push absences into the new year deliberately.
The rolling 12-month approach is generally cleaner: it captures the actual current pattern without giving a false reset on a fixed date.
Threshold trigger points
The standard policy translates score into action:
| Score | Typical action | |-------|----------------| | 51 | Informal conversation with line manager | | 201 | Verbal warning under absence policy | | 401 | Written warning | | 601 | Final written warning | | 801 | Review for dismissal on capability grounds |
These are guidelines from CIPD and HR consultancy practice, not statutory levels. The actual triggers should be set in your absence-management policy — and applied consistently.
The progression matters: Bradford Factor on its own does not justify dismissal. It identifies where attention is needed; the underlying capability or conduct case must be built independently.
What to exclude from the score
A naive Bradford Factor application — counting every absence — will quickly run into Equality Act trouble. Most well-designed policies exclude:
- Pregnancy-related absences — automatic discrimination if counted
- Disability-related absences where reasonable adjustments are in play — Equality Act risk
- Absences linked to a known long-term condition with appropriate medical certification
- Hospital appointments — typically time off, not full sickness absence
- Bereavement leave — usually a separate category
- Quarantine or public health-mandated absences — covered separately
Without these exclusions, the Bradford Factor becomes a backdoor way to discipline workers for conditions they cannot control. Tribunals see this; document the exclusions in policy and apply them consistently.
The legal risks
Disability discrimination
The biggest risk. A worker with a chronic condition (migraine, IBS, mental health, autoimmune) often has a pattern of frequent short absences — exactly what Bradford penalises. Section 20 of the Equality Act 2010 requires reasonable adjustments to working practices that put disabled workers at a substantial disadvantage.
Reasonable adjustments to absence policy may include:
- Discounting disability-related absences from the count
- Raising the trigger threshold for that worker
- Treating each disability-related spell as one continuous spell rather than multiple
Failure to consider adjustments → liability for disability discrimination, including injury-to-feelings awards (£1,200–£59,300+ depending on band).
Indirect discrimination
A neutral rule that disproportionately disadvantages a protected group can be indirect discrimination. The Bradford Factor's structure can disadvantage:
- Women with caring responsibilities (more likely to have child-illness short absences)
- Older workers (more likely to have age-related health issues)
- Workers with mental health conditions
The defence is objective justification — showing the rule is a proportionate means of achieving a legitimate aim. That defence requires documentation, not assertion.
Constructive dismissal
A worker subjected to inappropriate Bradford-driven action may resign and claim constructive dismissal. With two years' service, this is an unfair dismissal claim with no qualifying period restriction.
How to use the Bradford Factor properly
The score should inform, not decide. A robust application:
- Calculate the score for the rolling 12-month period
- Trigger a review meeting at the first threshold (typically 51)
- Investigate the underlying cause — is there a pattern? A condition? A workplace issue?
- Apply exclusions for protected categories before any disciplinary action
- Consider reasonable adjustments for any disability-related component
- Document the meeting and outcomes
- Apply progressive action only where the underlying issue is conduct or capability, not health
The score is the starting point of a conversation, not the conclusion of one.
Variations and alternatives
Modified Bradford
Some employers use S × D (without squaring) to reduce the frequency penalty. This is sometimes called the "modified Bradford" or "weighted attendance score."
Lost-time rate
A simpler measure: total days lost ÷ total available days, expressed as a percentage. Doesn't capture pattern — but is harder to challenge as discriminatory.
Frequency rate
Counts spells of absence per worker per year. Pure frequency, no weighting.
Some employers use a dashboard of all three measures rather than relying on Bradford alone. This gives a fuller picture of pattern, total impact, and frequency separately.
What it doesn't measure
The Bradford Factor measures observable absence. It doesn't measure:
- Presenteeism — coming to work sick and being unproductive (or contagious)
- Underlying engagement — disengaged staff have many small absences for low-grade reasons
- Workplace causes — is the absence pattern driven by something in the work itself?
- Manager quality — some teams have high Bradford scores because of management issues, not worker issues
A manager looking only at the Bradford Factor will miss most of these — and miss the chance to address the actual cause.
Implementation pitfalls
"The score went up, so we acted"
The score is a flag, not a verdict. Acting purely on the number — without investigating the underlying pattern — is the most common Bradford failure mode.
Applying the same threshold to everyone
A worker with a known disability may need a higher threshold (or different policy entirely) as a reasonable adjustment. Uniform thresholds across the workforce create discrimination risk.
Counting absences twice
Linking rules under the SSP regulations may treat two short spells within 8 weeks as a single linked period. The Bradford Factor's "spell" count can use the same logic to avoid double-counting genuinely linked absences.
Manual calculation
Calculating Bradford by hand on a spreadsheet leads to errors. Use HR software that calculates it automatically over a rolling window.
Putting it into practice
A robust Bradford Factor implementation:
- Defines the formula and reference period in policy
- Sets clear threshold-action mappings
- Documents the categories of absence excluded from the calculation
- Calculates the score automatically across a rolling 12 months
- Triggers manager-led review meetings, not automatic disciplinary action
- Builds in a reasonable-adjustment review for any worker approaching a threshold
- Reports patterns at team / department level for management insight
The Bradford Factor remains useful precisely because frequent short absences are more disruptive than rare long ones — that observation is not wrong. The error is treating the score as a verdict rather than a flag, and applying it without the safeguards that protect against discrimination claims.
Frequently asked questions
- How do you calculate the Bradford Factor?
- Multiply the number of separate spells of absence (S) squared, by the total days absent (D). For example, four 1-day absences gives 4² × 4 = 64. One 4-day absence gives 1² × 4 = 4. Same total days, very different scores.
- What time period does the calculation cover?
- Most employers use a rolling 12-month period. The score recalculates as the rolling window moves, so old absences fall off and new ones come on.
- What scores trigger action?
- Common thresholds: 51 — informal conversation; 201 — verbal warning; 401 — written warning; 601 — final written warning; 801 — review for dismissal. These are guidelines — your absence policy should set the actual triggers.
- Should every absence count towards the score?
- No. Most policies exclude pregnancy-related absence, disability-related absence (where adjustments apply), and absence covered by a doctor's note for a long-term condition. Including these can amount to discrimination.
- What are the criticisms of the Bradford Factor?
- It can punish workers with chronic conditions or genuine intermittent illness; it discourages people from staying home when sick; and it scores all short absences identically regardless of cause. It works best alongside qualitative review, not as the sole trigger.