The premium the payslip does not show

Overtime has a list price everyone knows — the premium: +25%, +30%, +50% depending on contract, schedule and country. And it is a price that lies low, because it compares pay, not production: the extra hour is paid more and yields less. The ninth and tenth hours of a day do not produce what the third does — fatigue research has measured it for a century, every floor supervisor has always known it.

The honest count corrects the premium for the yield: at the reference — a 25% premium, 85% marginal productivity — the effective multiplier is 1.47: the «+25%» on the payslip is +47% in production. Every overtime hour that enters a quote at the contract rate is giving away the difference.

This calculator gives the monthly cost of the overtime pool, the effective multiplier, the cost per hour actually produced — and the number that changes decisions: how many equivalent people that pool of hours represents.

The formula: premium plus fatigue

The effective multiplier divides the cost of the uplifted hour by the yield of the tired one: (100 + premium) / marginal productivity. It is a dimensionless number — valid in every currency and contract — and it has an uncomfortable property: the two components multiply. The premium raises the numerator, fatigue lowers the denominator, and the result grows faster than either.

Marginal productivity is the input no ERP exposes but everyone knows roughly: how much the extra hour yields versus an ordinary one. The fatigue literature places the ninth-tenth hour's yield between 70 and 90% depending on the work — physical or cognitive, repetitive or varied — and lower still when overtime is chronic, because tiredness accumulates across days, not only within the day.

The starting hourly cost must be the full one — pay plus contributions and charges — because in most jurisdictions the premium drags the charges along with it: the uplift applies to the employer cost, not to the paycheck alone.

The same overtime, five contracts

The table holds the extra hour's yield fixed — 85% — and the full ordinary-hour cost (28), and moves only the premium: the same tired hour, five list prices.

The middle column is the one to bring to the meeting: even the mildest premium in the table costs a third more than the ordinary hour, and the +75% night-or-holiday one exceeds double. The spread between the list and the effective grows along the whole scale — because fatigue does not read the contract: it discounts every row the same way.

The table is also the price list for whoever sells hours: the urgent job delivered on +50% overtime costs 49.41 in production against the 28 of the planned hour — a 76% premium the «urgency» quote should contain in full, and almost never does.

How this calculator works

The full hourly cost is the company cost of the ordinary hour: pay plus contributions and charges — the paycheck alone understates by a third and more. The premium is your contract's or practice's percentage uplift, with higher values for nights and holidays.

The monthly hours per person are the real average of those who work overtime — from the timesheets, not the contractual cap — and the people involved are how many do it regularly, not the whole workforce. The marginal productivity is the extra hour's yield versus an ordinary one: 85% is a reasonable default for mixed work, to be lowered for chronic overtime or heavy tasks.

From here come the monthly cost, the effective multiplier, the cost per produced hour and the workforce equivalent — on the conventional 174-hour full-time month, declared and constant.

How to read the results

The monthly cost is the income statement item: what the overtime pool really weighs, month after month. The cost per effective hour is the number for quotes: what an hour of production bought as overtime costs, to compare with the planned hour, the extra shift and the outside.

The effective multiplier is the verdict: below 1.25 the premium is contained and occasional overtime is the right tool for peaks; between 1.25 and 1.5 the hidden premium is already significant — the number to use in the accounts is this one, not the list; between 1.5 and 1.75 it sustains only high-margin work or emergencies; beyond 1.75 it is the most expensive way of buying hours the company knows.

The ghost employees

The fourth output is the one that moves decisions: at the reference, 12 hours a month for 40 people make 480 hours — 2.8 full-time equivalents. Somewhere in the org chart, almost three people work who do not exist: no desk, no interview, no approved budget — only timesheet lines spread across forty payslips.

The right question is not «can we afford the overtime?» but «did we decide to staff three positions this way?» — because at a 1.47 multiplier those three ghost people cost almost half again as much as three real ones, not counting what the model does not see: the fatigue that accumulates, the evening errors, the turnover of those who cannot keep the pace.

The comparison with hiring has its honest counterweight though: the real person is a fixed cost that stays when the peak passes. The practical threshold is persistence — overtime that has lasted more than six months at the same level is no longer a peak: it is the proof the load is structural, and the ghost is worth hiring.

The desk neighbours

The full hourly cost this page asks for has a dedicated calculator: the true cost of an employee builds the company hour — gross pay, contributions, benefits, workstation — and it is the mandatory starting point: whoever starts from the paycheck alone underestimates the overtime too.

The other neighbour closes the decision loop: revenue per employee computes the hiring headroom — how many people the income statement has already financed. If this page's ghost equivalent is smaller than that page's hiring headroom, the company is already paying for people it has not yet hired: the easiest business case in history.

The sequence for the budget: overtime pool and effective multiplier (this page), full cost of the alternative hires, hiring headroom at target — three numbers, and the «overtime versus headcount» discussion stops being an arm-wrestle between departments.

Where the numbers come from

The hours come from the timesheets or the attendance system: the average of those working overtime, not the average diluted over the whole workforce — and the distribution matters as much as the mean, because twenty people at 5 hours are a physiological peak, five people at 20 hours are a problem with a first and last name.

The full cost comes from the annual company cost divided by worked hours. The marginal productivity is estimated with honesty more than precision: for physical, repetitive work the literature indicates steeper drops; for cognitive work the drop comes later but sharper — and for both, chronic overtime yields less than occasional, because the missed recovery accumulates.

The practical advice: recompute at every contract renewal and every change of composition — the same premium on a doubled pool of hours is not the same problem doubled: it is a different problem, because marginal productivity falls with volume.

Limits of the estimate

Marginal productivity is a declared estimate, not a measurement: the true value varies by task, person and chronicity, and the calculator treats it as constant. The direction of the error is known though: whoever uses 100% — that is, ignores fatigue — always underestimates, and by a lot.

The model does not count the second-order costs: injuries and errors that grow with the hours, turnover of those who cannot keep up, recovery absenteeism — all worsen the picture in the same direction. And it does not check the legal limits: hour caps, mandatory rests and local rules stay outside the count and above it.

A parametric estimate for information purposes: it supports budgets, quotes and overtime-versus-headcount comparisons; it replaces neither employment law advice on the applicable limits nor the direct measurement of productivity.

The marginal-productivity scale

The first table holds the reference premium fixed — +25% — and moves the extra hour's yield: the same contract, five levels of tiredness.

The first row is the list: the multiplier you read in the contract, true only if fatigue does not exist. The last is the warning: at a 65% yield the same +25% costs almost double the ordinary hour — and the yield falls exactly when overtime turns chronic, that is, when the pool of hours is biggest. The worst unit cost always arrives together with the worst volume.

The operational reading: marginal productivity is not a parameter to set once — it is a variable the company moves. Shift rotation, real breaks, overtime spread across more heads instead of concentrated on the usual ones: every point of yield recovered is worth more than a point of negotiated premium.

The ghost scale

The second table moves the total monthly pool of hours — from a physiological peak to a shadow department — at equal contract and yield.

The last row is a whole department that does not exist: ten equivalent people, 60,900 a month, zero interviews held. No board would approve ten hires without discussing them — but the pool of hours gets there one timesheet at a time, without ever passing through a decision.

It is structural overtime's most insidious trait: it grows below the attention threshold of any approval process. The antidote is administrative — the pool of hours in equivalent FTE inside the monthly report, next to the headcount: when the ghosts have a number, sooner or later someone asks whose they are.

Nights, holidays, on-call: the compound premiums

The premium scale does not end at +25%: nights, holidays and combinations reach +50, +75 and beyond — and fatigue works against you here too, harder: the night hour pays the highest premium exactly while physiology yields the least, because the body discounts the circadian rhythm on top of tiredness. The chronic night multiplier easily exceeds 2.

On-call is the limit case: the allowance looks small because it pays «unworked» hours — but the night call-out it produces combines maximum premium, minimum yield and a compromised next day. The true cost of a night intervention, counted to the end, is often triple the list.

For all these bands the mathematics of this page stays identical: the band's premium in the numerator, the band's honest yield in the denominator. The numbers change, the formula does not — and the comparison between bands finally becomes homogeneous.

The urgency quote

Whoever sells hours — workshops, firms, agencies, maintenance — meets the effective multiplier every time they accept an impossible deadline: urgent work is done on overtime, and the quote written at the ordinary rate gives away the entire hidden premium. At the reference: 41.18 of cost against 28 of rate — 47% of the delivered labour cost gifted along with the urgency.

The honest price list has two columns: the planned rate and the urgency rate, with the ratio between them never below the effective multiplier. It is not a penalty on the customer — it is the true cost of the hours the urgency consumes, and declaring it upfront transforms the negotiation: half the urgencies stop being urgent when they discover their own price.

The internal corollary: in-house projects buy urgent hours too. The sprint that «recovers» the delay through evenings is buying hours at 1.47 times the price — and the delay recovered that way costs more than the delay accepted, almost always.

Overtime, extra shift, external: three prices for the same hours

Above a certain persistence, overtime competes with two alternatives. The extra shift or part-time: ordinary hours with no premium and full yield — it costs the onboarding and the contract's rigidity, but at equal hours it beats the multiplier almost always; it is the right move when the load is stable but not enough for a full-timer.

The external — agency, subcontractor, freelancer: an hourly rate above the ordinary internal one but often below the effective cost of overtime, with the advantage that it switches off when the peak passes. The right comparison is the external rate against this page's cost per effective hour, not against the internal list.

The practical hierarchy: overtime for short, unexpected peaks; external for long but temporary peaks; headcount for the load the timesheets have declared permanent for two semesters. Every skipped step is paid with the multiplier.

Mistakes that distort the count

The first is the paycheck instead of the full cost: the premium applies to the employer cost — contributions included — and starting from net or simple gross understates the whole pool by a third. The second is the diluted average: dividing the hours over the whole workforce instead of over those who work them hides the five people at 20 hours behind twenty people at zero.

The third is the 100% yield: ignoring fatigue is the implicit choice of anyone who has never seen this formula — and it always underestimates, never the opposite. The fourth is the contractual cap instead of the timesheets: the real pool includes unused compensatory rest and the «implicit» overtime nobody records, and the underestimate here adds to the others.

The fifth is strategic: treating the pool as a single number. Eighty hours spread over forty people and eighty hours concentrated on four are two different problems — the first is a cost, the second is a risk with a cost inside. The distribution must be looked at before the average.

Frequently asked questions

What is the true cost of an overtime hour?

The full cost of the ordinary hour times the effective multiplier: (100 + premium) divided by the extra hour's marginal productivity. At the reference — 28 full cost, +25% premium, 85% yield — the overtime hour costs 41.18 per hour actually produced: a +47%, not a +25%.

Why does the +25% on the payslip become +47% in production?

Because the two components multiply: the premium raises the hour's price (×1.25) and fatigue lowers its yield (÷0.85). The effective multiplier 1.25/0.85 = 1.47 compares what really matters — the cost per unit of production, not per hour of presence.

What is the marginal productivity of the extra hour?

How much the ninth or tenth hour yields versus an ordinary one. The fatigue literature places it between 70 and 90% depending on the work, and lower when overtime is chronic. The 85% default is reasonable for mixed work; 100% — that is, ignoring fatigue — always underestimates the cost.

Which premium should I enter?

Your contract's or practice's for the band of hours you are evaluating: base premiums typically sit between 25 and 50%, with higher values for night and holiday work. The calculator accepts 0 to 300% precisely to cover the extreme cases.

Why use the full hourly cost and not the paycheck?

Because in most jurisdictions the premium drags contributions and charges along with it: the uplift applies to the employer cost. Starting from pay alone understates both the ordinary hour and, in cascade, the whole overtime pool.

What is the workforce equivalent (FTE)?

The monthly pool of hours divided by the 174 hours of the conventional full-time month: at the reference, 480 hours = 2.8 equivalent people. It is the «ghost employees» number: positions staffed by overtime that nobody ever deliberated.

When is overtime "structural"?

The calculator's practical threshold is 16 monthly hours each — four a week, every week. At that level it does not cover peaks: it covers a missing headcount disguised as flexibility, and the comparison with hiring must be made in writing. The other signal is persistence: six months at the same level is no longer a peak.

Is hiring better than paying overtime?

It depends on multiplier and persistence: at 1.47× the ghost people cost almost half again as much as real ones, but the real person is a fixed cost that stays when the peak passes. The practical rule: occasional overtime for peaks, hiring when the volume has been stable for semesters.

Why does the calculator show no currency?

Because the mathematics is identical in every country: enter the hourly cost in your currency and the costs come out in the same unit, while the effective multiplier is a pure number comparable across contracts, departments and countries.

Does the model work for cognitive work too?

Yes, with a different marginal productivity: in knowledge work the yield drop comes later in the day but steeper, and the evening hours' errors cost rework the model does not count. For cognitive roles a value between 70 and 80% is often more honest than 85%.

What does this calculation not include?

The second-order costs: injuries and errors that grow with the hours, turnover of those who cannot keep the pace, recovery absenteeism. All push in the same direction — the true cost of chronic overtime is higher than shown.

Does the calculator check the legal limits on hours?

No: hour caps, mandatory rests and local rules stay outside the count and above it. This page measures the economic cost; regulatory compliance must be checked against the rules applicable to your contract and country.

Updated July 2026