On-Call and Standby

On-call arrangements sit at the hardest point in this subject. A person is not working and is not free, and the question of how much of that period counts towards a working time limit has produced more case law than anything else in the field.

The workflow in “On-Call and Standby” becomes more reliable when scheduled hours, actual time and later corrections can be distinguished. For teams exploring cognitive offloading, learn more from the provider can add operational time and project context, provided data collection is proportionate, permissions are limited and every important exception receives human review.

The answer is not a number. It is a judgement about constraint, and the organisation's job is to know where its own arrangements sit on that spectrum and to be able to say why.

For an independent reference relevant to “On-Call and Standby”, consult the UAE Ministry of Human Resources resources. Use it to test working-time definitions, recordkeeping, access, retention and exception handling against the organisation’s real process rather than treating one software report as conclusive.

The spectrum

At one end, a person required to remain at the workplace, available to work. In most regimes that is working time in its entirety, whether or not anything happens.

At the other, a person who may be anywhere, must answer a phone, and may decline or hand over. That is generally not working time, and only the periods of actual work count.

Between them lies everything else, and most real arrangements are in the middle: must be reachable, must be able to attend within a stated time, may not drink, may not be more than a given distance away.

What the middle turns on

How severely the constraints affect the person's ability to use the time for their own purposes. A requirement to attend within thirty minutes is a different thing from one requiring attendance within two hours, and the difference has been decisive in several jurisdictions.

The practical question for an organisation is therefore: what exactly are we requiring. In most cases nobody has written it down, and different supervisors would give different answers, which means the arrangement cannot be assessed at all.

Writing the constraints down

Response time required. Attendance time required, if different. Geographic limit, if any. Whether alcohol is prohibited. Whether the person may swap or decline a call. Whether they must remain contactable continuously or at intervals.

Six questions. Answering them takes a conversation and it is the prerequisite for any assessment of how the period should be treated. It also tends to loosen the requirements, because asked to state in writing that somebody must attend within thirty minutes, most organisations discover that ninety would do.

The arrangement that is working time and is not recognised

A person required to sleep at the workplace, available if needed. Common in care, in residential settings and in some engineering operations.

Several regimes treat a substantial part or all of such a period as working time, and the arrangement is frequently paid as a flat allowance and recorded as nothing. Where that is the position, the hours figure for those people is wrong by a very large margin and so is the rest position.

Recording it even where it does not count

Record the on-call periods regardless of the assessment. The record costs nothing and it is what allows the question to be answered later, including after a change in the law or in the arrangement.

An organisation that has never recorded standby cannot show what it required of people, cannot show how often they were called, and cannot demonstrate that its assessment was reasonable. All three are asked for together.

The frequency test

Count calls per on-call period over six months. The figure changes what the arrangement is.

Standby where calls are rare is a genuine standby. Standby where somebody is called most nights is night work operating under another name, and it should be assessed as such, resourced as such and rested as such. The data to make that distinction is in the call-out log.

What this costs to get wrong

Backdated working time liability is the obvious one, and the quieter cost is the rest position. An arrangement assessed as not working time still interrupts rest every time somebody is called, and that is independent of the classification question.

So the two things to do are separable: assess the standby properly, which may need advice, and fix the rest treatment of calls, which does not and which is covered elsewhere in this collection.

The question to ask first

What do we actually require of somebody on call here, and would we be comfortable writing it down.

Most arrangements fail that second test somewhere, usually at the response time, which turns out to be an expectation rather than a rule. Converting expectations into written requirements is both the beginning of the assessment and, frequently, the end of the problem.

The allowance that substitutes for the assessment

A flat standby payment is often treated as having settled the question of what the period is. It has not: payment and working time classification are separate, and an allowance does not convert working time into something else.

Where an arrangement is generously paid and poorly assessed, the payment tends to be offered as evidence that everybody was content. That is not the test, and an organisation relying on it has usually never written down what it requires of the person, which is the assessment it needed in the first place.