BAMML Executive Insights
The Overtime You Approved Last Month Is Next Year's Safety Incident
Overtime arrives at the executive table as a number. It is approved, month after month, as a cost line — sometimes reluctantly, usually because production has to be made up and the alternative is missing the plan. The approval takes a few minutes.
What is approved in that meeting is not only a cost. It is a fatigue input into a workplace where people operate machinery underground and at height, and where the consequences of a lapse in concentration are not measured in rands.
The difficulty is that the two arrive on different desks, months apart.
What is actually happening
On most operations, overtime is tracked in three places by three functions, and reconciled in none of them.
Finance sees it as payroll variance. The question asked is whether the overtime bill is within budget, and the answer is usually a percentage against forecast.
HR sees it as hours. The question is whether the hours comply with the applicable limits and agreements, and whether leave and rest periods are being honoured.
Operations sees it as availability. The question is whether there are enough bodies on shift to hit the tonnage.
Nobody sees it as an accumulating physiological load on a specific group of people doing a specific set of tasks. That view does not exist in any of the three systems, because it does not belong to any of them.
Why it matters
Fatigue has a lag. It builds across weeks of extended shifts, shortened rest periods and consecutive days worked. The incident, when it comes, does not announce itself as a fatigue incident. It is recorded as an error of judgement, a procedural failure, a slip, a near-miss that became something else.
By then the overtime that contributed to it is historical. It has already been approved, paid, reported and closed out. Nobody goes back to the March payroll variance when investigating a September incident, because there is no mechanism that would prompt anyone to do so.
This is why overtime, absenteeism, sick leave and safety incidents are so often managed as four separate problems by four separate people. Each has an owner. The pattern that connects them has none.
And the pattern is usually circular. Sustained overtime produces fatigue. Fatigue produces absence and sick leave. Absence pushes the remaining workforce into more overtime to cover the gap. The cost of that overtime squeezes the budget for supervision, training and maintenance — which are precisely the things that would have caught the next incident.
The Integrated View
Read across the five disciplines, the same decision looks different in each:
| Discipline | What overtime looks like here | The question nobody asks |
|---|---|---|
| Workforce Governance | Hours, rest periods, agreements, IR exposure | Which specific crews are carrying the load, and for how many consecutive cycles? |
| Financial Governance | Payroll variance against budget | What is overtime costing us once incident, absence and turnover costs are added back? |
| Operational Performance | Shift cover and tonnage | Are we hitting tonnage because of overtime, or despite the fatigue it is creating? |
| Risk & Compliance | Statutory duty to provide a safe working environment | Could we evidence that fatigue was assessed as a risk, and controlled? |
| Executive Decision Intelligence | Four separate reports | Is there one view that shows these moving together? |
The last row is the real finding. In most operations the answer is no — and not because anyone decided it should be that way.
What changed for one operation
One BAMML client, an industrial operation, carried exactly this cluster: high overtime, high absenteeism, high sick-leave records, workplace accidents and incidents of fuel theft. For years these were treated as unrelated problems. Together they were holding tonnages below target, pushing the payroll burden up and pulling financial targets down.
What changed was not new policy. It was aligning individual and team performance to the organisation's actual targets, and enforcing full, consistent implementation of the policies that already existed on paper.
The result: staff turnover fell below 2%, overtime reduced to public holidays only, and the operation went more than a year with no reported accidents.
Notice what that last sentence contains. The safety outcome was not produced by a safety intervention. It was produced by fixing the workforce governance underneath it.
What to do on Monday
Ask for one report that does not currently exist: overtime hours, absenteeism, sick leave, disciplinary cases and safety incidents, for the last twelve months, broken down by crew and by supervisor — on one page.
Then work through five questions:
- Which crews carry the overtime? If it concentrates in the same teams, this is a structural staffing problem, not a demand problem.
- Do absence rates rise in the weeks following heavy overtime periods? If they do, you are already paying twice.
- Can we evidence that fatigue was assessed as a risk and controlled? Not whether you believe it is managed — whether you could produce the record.
- What is the real cost of an hour of overtime? The rate, plus the absence it produces, plus the turnover it contributes to, plus the incident exposure it creates.
- Who owns the pattern? If the answer is that four people each own a piece, nobody owns it.
If you cannot produce that single page, that is the finding — and it is a governance finding, not a reporting inconvenience.
