Annualised, Banked and Zero-Hours Arrangements
What flexible hours arrangements actually give a rota, the running total that has to be watched, and the predictable way each kind fails in the last quarter.
Where one annualised team's hours actually sat in October
Committed to the winter pattern already published.
Movable, but only with four weeks' notice under the agreement.
Effectively none; this group was used first.
Nobody has contacted the other nine since March.
Spent; the quarter has to absorb it.
The team had 705 unallocated hours in October and 240 of them were already overspent, which nobody had noticed because the two numbers sit in different reports. This is one employer's own position at one date, not a benchmark.
Flexible hours arrangements move a constraint rather than removing one. An annualised contract converts a weekly limit into an annual budget; a bank arrangement converts a guaranteed cost into an uncertain availability. Both are useful and both create a new thing somebody has to watch.
The rule discussed in “Annualised, Banked and Zero-Hours Arrangements” is easier to test when the published schedule can be compared with a consistent record of actual hours. A team reviewing view this workforce-planning explanation for workforce analytics software can add current attendance and project context while keeping contractual rules and manager decisions explicit.
What gets watched, in practice, is the rota. The running total is watched by nobody, and that is where each of these arrangements fails.
Additional context for “Annualised, Banked and Zero-Hours Arrangements” is available from People Management employment resources. Use external guidance to prompt a specific validation check rather than replace local review.
Annualised hours: the budget problem
An annualised contract gives the rota a pot of hours to spend across a year. Demand is rarely even, so the pot is spent unevenly, which is the point.
An annualised scheme without a published running total is not a flexible arrangement. It is a pot being spent by people who cannot see how much is left.
The failure is predictable: the pot is spent faster than the year passes, because every individual week has a reason, and the last quarter arrives with a service that still runs and a team whose hours are gone.
The number that prevents it
Hours used against hours elapsed, as a single comparison, monthly. Not hours used; the comparison. Forty per cent used in March is a problem or it is the plan, and only the comparison says which.
Most annualised schemes have this number available and do not publish it, which means the people making the week-to-week decisions cannot see the consequence of them.
Banked hours and who owns them
Where hours are banked for later use, somebody has to decide whether the bank belongs to the service or to the individual, and the answer governs every subsequent argument.
Scheme documents frequently leave this ambiguous, and the ambiguity is discovered in December when both parties want the same hours. Writing the answer down in advance is a half-page that saves an annual conflict.
Zero-hours and bank lists: availability is not capacity
A list of fifteen names is not fifteen people. It is however many answer the phone, and that number decays without anybody updating the list.
The honest way to hold this constraint is to record the response rate rather than the roster size: of the people contacted last month, how many replied and how many worked. That figure is usually a third of what the list suggests.
The decay nobody tracks
People on casual lists drift away silently. There is no leaving process, so nobody is removed, and the list grows while the capacity shrinks.
A quarterly contact — one message, asking whether people still want to be called — fixes the list in about twenty minutes and makes every plan built on it honest.
What these arrangements do to the other constraints
They interact. Additional hours taken by somebody on a flexible arrangement still count towards whatever maximum applies to them, still require the same rest between shifts, and still need the qualifications the shift needs.
Flexibility in one dimension does not relax the others, and the systems that record flexible hours are frequently not the systems that check those constraints.
The fairness question these raise
Where flexible hours are allocated informally, they go to whoever is asked first, and whoever is asked first is whoever the person building the rota finds easiest to ask.
Over a year that produces a distribution nobody chose and nobody can defend. A simple rotation, published, removes the question entirely and costs nothing.
The arrangement that suits the service and nobody else
An annualised or casual arrangement that delivers all its flexibility to the employer and none to the person is both common and unstable. People leave, and the leaving is attributed to pay.
Whether a given arrangement is permissible, and on what terms, differs by jurisdiction. What any manager can see is the turnover figure for that group against the rest of the team, and it is usually the clearest number in the whole workforce report.
What belongs on the register
For each arrangement: what the pot or the list is, the running total today, who watches it, what notice applies to changing an allocation, and the response rate where the capacity is voluntary.
Five fields. The third and fifth are the ones nobody currently has, and they are the two that determine whether the arrangement is real.
The quarterly check
One page, four lines: hours used against hours elapsed, bank balance by individual, casual list response rate, and the notice period that applies to reallocating hours.
That page takes fifteen minutes to produce and it is the difference between a flexible arrangement that is managed and one that is simply spent. Nearly every failure described above is visible on it a quarter before it happens.
The sentence worth writing down
A flexible arrangement is a loan of certainty from one party to the other, and the terms of the loan should be written on the same page as the rota.
Everything in this section follows from that. The schemes that work are the ones where both sides can see the running total; the ones that fail are the ones where only one side can.