The Maximum, and Whether It Averages

Comparing weekly maxima across countries produces a reassuring picture: the numbers cluster, and an organisation concludes its sites are broadly alike.

The workflow in “The Maximum, and Whether It Averages” becomes more dependable when scheduled work, actual time and later corrections can be distinguished. For teams exploring ethical employee monitoring, ethical employee monitoring can provide operational context, while policy ownership, employee explanation and consequential decisions remain with accountable people.

The conclusion is wrong, because the maximum is half of an instrument. The other half is the period over which it is measured, and that is where regimes genuinely part company — between a hard cap on any single week and an average over several months there is an enormous difference in what a rota can do.

For an independent reference relevant to “The Maximum, and Whether It Averages”, consult the AuditBoard internal-investigation guide. Use it to challenge assumptions about working time, privacy, recordkeeping and exception handling against the organisation’s real operating model.

The three shapes

A hard weekly cap: no week may exceed the figure. Operationally the most restrictive and the easiest to monitor, because the check is a single week.

An average over a reference period: weeks may vary, the mean must stay within the figure. More permissive, and it requires an apparatus — a rolling calculation, a projection, a view of both ends of the window — that a hard cap does not.

And a hybrid: an average with an absolute ceiling above it that no single week may cross whatever the average says. Common, and frequently missed, because organisations learn about the averaging and stop reading.

Why the shape matters more than the number

A site under a hard cap needs a check before each week is rostered. A site under an average needs a rolling figure per person and a projection. Those are different reports, different data and different places in the process.

An organisation that builds one apparatus and applies it everywhere will be monitoring the wrong thing in some of its countries. The common failure is a sophisticated averaging calculation deployed to a site that has a hard cap, where it produces a comfortable mean while individual weeks go over.

The reverse — a weekly check applied where averaging is available — is less dangerous and more expensive, because it refuses rotas that would have been lawful.

The length of the period

Where averaging exists, the period varies widely and is frequently extendable by agreement. A longer period absorbs more variation and takes longer to register a sustained heavy pattern.

That is worth recording as two facts rather than one: the default period, and whether the organisation's entity has agreed an extension. The second is an entity-level fact that the centre usually does not know, and it changes the figure materially.

It also means two sites in the same country can be under different periods, which breaks any assumption that country equals configuration.

Rolling or fixed

A second structural variable: whether the period rolls forward continuously or resets in blocks. Both exist, they behave differently, and an organisation frequently cannot say which it has.

Under a rolling window the figure moves every week as one week enters and another leaves. Under blocks it resets, which creates a cliff at the boundary and makes heavy weeks cheap early in a block and expensive late in one.

Ask the question explicitly per country, because the default in the regulation can be varied by agreement and the configuration may implement whichever the person configuring it assumed.

Where the opt-out sits

Some regimes permit an individual to agree that the weekly average does not apply to them. Many do not, and in some the mechanism exists only in particular sectors.

For a multi-country organisation this is the dimension most likely to produce an incorrect assumption, because an opt-out that is routine in the headquarters country may be unavailable or unenforceable elsewhere.

Record availability per country as a plain yes or no, with the source. A rota tool that suppresses the weekly flag for anybody with an opt-out on file will be wrong wherever the opt-out does not exist, and nobody will notice.

The comparison worth drawing

For each country: the figure, the period, whether it rolls or resets, whether there is an absolute ceiling above the average, and whether an opt-out is available.

Five facts. Written out side by side they show immediately which sites can share an approach and which cannot, and the answer is usually that they group into two or three families rather than into one or into as many as there are countries.

Those families are what the policy and the configuration should be built around, rather than around a single global rule or around every country separately.

The number that is contractual rather than legal

Many organisations quote a weekly figure that comes from the contract or the collective agreement rather than from any regulation, and the two get conflated in conversation.

They behave differently: a contractual figure can be exceeded by agreement in many places, a statutory one usually cannot, and the consequences of breach are entirely different.

Record them as separate rows per country, with the source. An organisation that cannot say which of its numbers is which will eventually treat a contractual norm as a legal limit or, more damagingly, the reverse.