The Bradford Factor in South Africa: A Practical Guide for HR Teams

Where the Bradford Factor Comes From

The Bradford Factor is usually attributed to research at the University of Bradford School of Management in the United Kingdom in the 1980s. We could not find the original study, so treat that origin as an attribution rather than an established fact. The idea behind it is one many line managers already hold: frequent short absences are harder to manage than a single longer absence of the same total length.

An employee absent for twelve days in a single stretch is administratively manageable. The organisation can plan around it, arrange cover, communicate with clients, and prepare for the return. An employee absent for twelve days scattered across six separate one-day events throughout the year is a fundamentally different operational problem. Coverage cannot be pre-arranged. The absence is unpredictable. Each event generates a separate management response. The cumulative disruption is far greater than the cumulative day count suggests.

The Bradford Factor turns this idea into a number that ranks employees by how often they are absent.

The Formula

The Bradford Factor is calculated as:

Bradford Score = S² × D

Where S is the total number of separate absence spells in the measurement period, and D is the total number of days absent in the same period. The measurement period is typically a rolling 52 weeks, though some organisations apply it over a shorter 26-week window for more responsive early warning.

The squaring of S is what gives the formula its diagnostic power. Increasing the number of separate incidents compounds the score dramatically, while increasing only the duration of a single absence has a much more modest effect. This mathematical structure directly encodes the operational insight that frequency of disruption matters more than aggregate duration.

Reading the Numbers: What Different Scores Mean

Employers interpret Bradford Scores against threshold bands that they set themselves. There is no official or South African standard. The bands below are an example of how an employer might set them:

• 0 to 49: Low risk. No intervention required beyond routine leave management.
• 50 to 99: Caution. The pattern warrants monitoring and an informal conversation with the employee. No formal action is indicated unless the score continues to rise.
• 100 to 199: Elevated risk. A formal counselling session or documented verbal warning is appropriate, depending on the employee's leave history and any underlying health factors disclosed. This is the threshold at which most organisations begin a structured intervention process.
• 200 to 399: High risk. A written warning is typically warranted. HR should review the employee's full absence record and consider whether an incapacity or misconduct investigation is appropriate.
• 400 and above: Critical. A full review of the absence record is needed. The cause may be a chronic illness, which calls for incapacity management, or a conduct problem, which may call for the misconduct process. The score alone does not tell you which.

These thresholds are guidelines, not law. Your organisation's policy should define the trigger points and the corresponding management responses, and those responses must be applied consistently across all employees. Inconsistent treatment weakens the fairness of any later dismissal (item 10 of the Code of Good Practice: Dismissal (2025)).

Worked Examples: The Formula in Practice

The best way to understand why the Bradford Factor is diagnostically superior to a simple absenteeism rate is to compare employees who look similar on a raw day-count basis but produce very different Bradford Scores.

Employee A: Absent for 12 days in total over the past year, in a single continuous spell following a hospitalisation. S = 1, D = 12. Bradford Score = 1² × 12 = 12. Low risk. This is consistent with a genuine health event and presents minimal operational disruption.

Employee B: Absent for 12 days in total over the past year, across four separate spells of three days each. S = 4, D = 12. Bradford Score = 4² × 12 = 192. Elevated risk. Four separate events have generated four separate coverage problems, four separate management responses, and four separate periods of team disruption.

Employee C: Absent for 12 days in total over the past year, across twelve separate one-day absences. S = 12, D = 12. Bradford Score = 12² × 12 = 1,728. Critical. Twelve separate absences are much harder to cover than one long one. A run of one-day absences is a reason for a conversation with the employee. It is not proof that the sick leave was not genuine: some conditions, such as migraine or a chronic illness with flare-ups, produce exactly this pattern.

All three employees have taken the same number of days off. A simple absenteeism rate treats them identically. The Bradford Factor correctly identifies that Employee C presents a fundamentally different management challenge to Employee A.

Why Frequency Matters More Than Duration in the South African Context

In the South African labour market, where a significant portion of the workforce is employed in operational roles that require physical presence, a single unplanned one-day absence triggers a chain of operational consequences that a planned extended absence does not. In healthcare, logistics, manufacturing, retail, and security, coverage for a single shift requires a replacement to be sourced, briefed, and deployed, often at overtime rates, within hours. The administrative and cost overhead per event is largely fixed regardless of the length of the absence that triggered it.

In professional services and knowledge work environments, the calculus is slightly different, but the frequency problem remains. A consultant absent for a single day in the middle of a project engagement may cause a client meeting to be rescheduled, a deliverable deadline to be missed, or a team dependency to be broken. The disruption from twelve such events over the course of a year is qualitatively and quantitatively worse than one planned leave period of the same aggregate duration.

The Bradford Factor captures this reality mathematically in a way that a simple "days absent this year" metric does not.

The Bradford Factor and South African Labour Law

The Bradford Factor is a management tool, not a legal standard. It does not appear in the Basic Conditions of Employment Act, the Labour Relations Act, or any sectoral determination. Its value in the South African context is as a structured, consistently applied trigger for management action, not as a standalone basis for dismissal.

This distinction matters enormously at the CCMA. An employer who dismisses an employee primarily on the basis of a Bradford Score, without conducting a fair investigation into the underlying cause of the absences, without applying the appropriate procedure (incapacity or misconduct depending on the nature of the absence), and without being able to demonstrate that the threshold and response were applied consistently to all employees, is exposed to an unfair dismissal finding regardless of how high the score is.

The Bradford Factor is a way to decide whom to look at, not a substitute for a fair process. For example, an employer might set its policy so that a score above 100 leads to a counselling conversation, a score above 200 to an investigation into the cause of the absences, and a score above 400 to a formal process. These cut-offs are examples set by the employer, not legal thresholds. The formal process would be incapacity management under item 21 of the Code of Good Practice: Dismissal (2025), which replaced Schedule 8 of the LRA on 4 September 2025, if the absences are genuinely health-related, or the misconduct process (items 6 to 11 of the Code) if they are found to be without valid cause.

The score identifies whom to look at. The Code of Good Practice: Dismissal sets the steps you must follow after that.

Limitations of the Bradford Factor

The Bradford Factor is a powerful tool, but it has real limitations that experienced practitioners need to understand before building policy around it.

First, the formula does not differentiate between types of absence. An employee with a high Bradford Score driven by a serious chronic condition that is well-documented, disclosed, and managed responsibly is in a fundamentally different position to an employee whose score is driven by unexplained one-day absences on Mondays and Fridays. The score looks identical. The appropriate management response is entirely different.

Second, the Bradford Factor can be gamed by employees who are aware of it. An employee who knows that three one-day absences produce a much higher score than one three-day absence may adjust their absence pattern accordingly. This is a reason to track the raw absence data in parallel rather than relying solely on the Bradford Score.

Third, the formula does not account for the nature of the role. A Bradford Score of 150 for an employee in a back-office administrative role that can be covered by task redistribution represents a different operational risk to the same score for a specialist nurse, a safety-critical technician, or a client-facing relationship manager. Policy thresholds should account for role criticality, not apply a single set of cut-offs to the entire organisation.

Fourth, new employees will mechanically have lower Bradford Scores regardless of how problematic their absence behaviour is, simply because the measurement window is shorter. Organisations should apply the Bradford Factor alongside a frequency-per-period metric for employees who are still within their first six to twelve months of employment.

Calculating the Bradford Factor at Scale

For a small team, the Bradford Factor can be calculated manually in a spreadsheet. You need two data points per employee per period: the number of separate absence events (S) and the total days absent (D). The calculation is straightforward. The challenge is keeping it current.

The Bradford Factor is most useful as a rolling metric, recalculated regularly against a moving window. A score calculated once a year at the end of June misses the point almost entirely. An employee who generates twelve separate one-day absences in the first six months of the year, producing a Bradford Score of 1,728, should be identified and managed in month two or three, not in month seven when the damage has already been done.

As headcount grows, keeping a current rolling Bradford Score for every employee in a spreadsheet takes more HR time, because every new or corrected leave record changes the score. Where possible, calculate it from the same data source that holds your leave records, so it updates when the records change.

This is the difference between absenteeism management as a retrospective reporting exercise and absenteeism management as an early warning system. The Bradford Factor is only an early warning tool if it is calculated continuously, not annually.

Using the Bradford Factor as Part of a Broader Risk Framework

The Bradford Factor works best when it is one component in a broader attendance risk picture rather than the sole metric an organisation tracks.

An employee with a moderately elevated Bradford Score who also shows a clear pattern of Monday and Friday absences, three public holiday adjacency events in the past six months, and a 30-day trend of rising frequency is a materially different management problem to an employee with the same Bradford Score whose absences show no pattern correlation and are supported by consistent medical documentation.

The Bradford Factor quantifies the operational disruption caused by absence frequency. Pattern analysis identifies the likely cause. The combination of the two gives HR practitioners and line managers the information they need to make confident, well-grounded intervention decisions.

A practical approach is to use the Bradford Factor to rank employees by frequency, use pattern analysis to decide which cases need a conversation first, and then find out the cause from the employee before deciding between support, incapacity management and discipline.

Implementing the Bradford Factor in Your Organisation

If you are implementing the Bradford Factor for the first time, the process has four practical steps.

The first step is to decide on your measurement window and your threshold triggers. A rolling 52-week window is the most common choice. Your threshold levels should reflect your sector, your operational structure, and your capacity for management intervention. Document these decisions in your absenteeism policy before you begin applying them.

The second step is to calculate baseline scores for your current workforce using your existing leave data. This gives you an immediate picture of which employees are already in elevated or critical territory, and allows you to begin priority-based intervention without waiting for the system to accumulate new data.

The third step is to ensure that your line managers understand the formula and the thresholds, and know what management action each threshold triggers. The Bradford Factor is useless if it sits in an HR spreadsheet and is never connected to a management conversation.

The fourth step is to establish a recalculation cadence. Monthly is the minimum. Weekly is better. Nightly is best, and is achievable without manual effort if your leave data is managed in a system that supports it.

The Bradford Factor will not fix absenteeism on its own. Used with a written policy and applied the same way to everyone, it helps you decide which employees to speak to first.

This article is general information, not legal advice. Check the current Basic Conditions of Employment Act, the Code of Good Practice: Dismissal (2025), and any sectoral determination or bargaining council agreement that applies to you.