In any workplace where shifts have to be covered, the roster is more than a timetable. It is the clearest daily statement of how the business values people's time. Who gets the Saturday, who gets the long weekend, who is asked to cover when someone is sick: each of these decisions says something about fairness, whether or not anyone meant it to.
This piece argues, as a hypothesis, that fair rosters depend less on the good intentions of whoever builds them than on whether the system behind them is transparent. When people can see how a roster was made and what happens when someone is absent, they are far more likely to accept an outcome they would not have chosen. When they cannot, even a fair roster tends to be read as a favour or a slight.
Why absence turns rosters into a fairness test
A roster is easy to accept on a normal week. Fairness is tested when someone is off sick, on family responsibility leave, or simply not there. The gap has to be filled by someone, and that someone is usually chosen quickly, by whoever is managing the day.
Quick decisions made by one person with incomplete information tend to follow the path of least resistance. The colleague who always says yes is asked again. The person who is nearest or easiest to reach gets the call. The one who has been covering for months is not recognised as having done so. None of this is malicious, but over time it can produce a pattern in which some people carry far more of the unpredictability than others.
That pattern is rarely visible to the manager, because it was never recorded. It is very visible to the staff who live it.
Fairness is about process as well as outcome
Researchers in organisational behaviour distinguish between fairness of outcomes and fairness of the process used to reach them. Wikipedia's overview of organisational justice describes this distinction and the finding that people's perception of how decisions are made, including whether the process is consistent and open, can shape their reactions as much as the result.
This is a general finding from a wide literature rather than a measurement of South African workplaces, but the implication for rosters is intuitive. A person who gets an unwelcome shift under a clear rule, such as a published rotation, is more likely to accept it than a person who gets the same shift because a manager made a call that no one can see the reasoning for.
The dignity tax of an opaque roster is therefore not mainly about the shift. It is about being unable to tell whether you were treated as an equal. For related reflections on how systems can signal trust or distrust, see Mellorca's piece on when the system decides what you can do today.
What the law sets as the floor
In South Africa the Basic Conditions of Employment Act sets the minimum framework within which rosters must sit. Among other things, it regulates ordinary hours of work, overtime, meal intervals, daily and weekly rest periods, and work on Sundays and at night, and it requires employers to give employees written particulars of their employment, including their hours of work. Sectoral determinations and bargaining council agreements may add or vary rules for particular industries, so employers should confirm what applies to them, and the Department of Employment and Labour is the right place to check.
The law is a floor, not a definition of fairness. A roster can be lawful and still feel unfair, because the Act does not tell a manager who should cover a gap on a given Tuesday. That judgement is left to the business, which is why the system around it matters.
A comparable body of guidance exists in other countries. The UK's Acas, for example, publishes practical advice on working time and rest, which illustrates how widely accepted the principle is that workers need rest and that employers should manage it deliberately. It is not South African law and is mentioned only as a point of comparison.
What transparency looks like in practice
A transparent roster system is not complicated. It has a handful of features that make decisions visible and checkable.
- Published rules. A short statement of how shifts are allocated: rotation of weekends, how requests are handled, how much notice is given for changes, and how cover is arranged during absence.
- A shared, current roster. One version, visible to everyone affected, rather than versions on a manager's phone, a noticeboard and a WhatsApp group that disagree.
- A record of who covered what. A simple tally of weekend shifts, public holidays, last-minute call-ins and cover for absence, so that burdens can be seen and rebalanced.
- A fair order for cover requests. Rather than asking whoever answers first, work through a known order, such as the person who has covered least recently, and allow people to decline without penalty where practical.
- Advance notice of changes. The sooner people know, the more of their lives they can plan around the roster.
- A way to query or swap. People should be able to ask why they were rostered and to arrange swaps within clear limits.
Scheduling software can support all of this. Wikipedia's description of employee scheduling software says it typically tracks leave and sick time, alerts users to conflicts, and keeps data that can later be used for payroll or to analyse past activity. A simple shared spreadsheet can do part of the job. The tool matters less than whether the rules and the record are open to the people affected.
The coverage side of the problem is discussed in Mellorca's article on why leave requests turn into coverage puzzles, and the person who ends up as the fallback for everything is explored in the employee who needs permission for every small fix.
An illustrative scenario
Consider a hypothetical example, not a real business: a Bloemfontein private clinic reception and admissions team of eight people covering seven days a week. The practice manager builds the roster each month on a spreadsheet kept on her own laptop and adjusts it through the month by phone.
When a receptionist is off sick, the manager rings the people she thinks will say yes. Over a few months, two staff members absorb most of the extra Saturdays, while another has never been called. No one has complained openly, but two of the team have started to talk about the manager having favourites, and the staff who cover most have begun to refuse calls, which in turn makes the manager rely on them even more reluctantly.
The manager changes three things. She publishes a one-page set of rules. She moves the roster to a shared file that everyone can see, with a column for each person that counts weekends and call-in cover. And she adopts a simple order for last-minute cover, starting with whoever has done the least. Within a couple of months, the complaints stop, not because everyone is delighted with their shifts, but because the pattern is now visible and defensible.
The human dividend
The dividend is predictability: knowing in advance when you work, knowing how changes are handled, and knowing that the burden of absence is being shared by a rule rather than by whoever is easiest to ask. A secondary gain is dignity. Being given a hard shift under a rule one can read feels different from being given it by chance or by favour.
The trade-offs
- Rules reduce flexibility. A rotation that is fair on paper may not suit a particular person's situation, and managers lose some room to accommodate individual circumstances.
- Visibility cuts both ways. A public tally can expose uneven burdens, but it can also create comparison and disputes that did not exist before.
- Tracking creates data. Records of shifts and leave are personal information and should be kept and shared with that in mind, in line with POPIA.
- Software can harden unfairness. An automated roster built on poor rules applies them consistently, which is not the same as fairly.
- Fair is not identical. Some staff need regular hours for childcare or study; others want more shifts for the income. A transparent system can honour those differences, but it requires openness about them.
Why it matters
A roster is where a business's values meet people's calendars. Staff accept hard shifts when they trust the way they were decided, and lose trust when the way is hidden, however kind the intent. Making the system transparent will not remove every disagreement. It will make disagreements discussable, and it will replace guesswork about favouritism with something people can read, question and improve.
Sources and further reading
- Basic Conditions of Employment Act 75 of 1997 - South African Government (last verified 10 October 2026)
- Organizational justice - Wikipedia (last verified 10 October 2026)
- Employee scheduling software - Wikipedia (last verified 10 October 2026)
- Working time and rest - Acas (last verified 10 October 2026)
Related Mellorca reading
- When Nobody Can Say Which Number Is True
- The Customer Who Gets a Different Answer From Every Person
- When Sunday Evening Becomes Monday Morning