Overtime and shift premiums in Colombia: how they are calculated
Overtime and shift premiums in Colombia: how they are calculated
A badly calculated overtime payment almost never gets the percentage wrong: it gets the base wrong, forgets the night premium that nobody added, or pays Sunday as an ordinary day. The correct mechanics are always the same: take the value of the ordinary hour, multiply it by the factor that applies —1.25 for daytime overtime, 1.75 for night overtime, 2.00 for daytime Sunday work— and multiply by the number of hours. Here are the four types of hour, the limits you cannot cross, the effect on social contributions and benefits, and a worked example you can replicate with your own payroll.
What counts as overtime and what does not
Overtime is work performed outside the agreed shift, with the company's authorization. It is not "any hour past six in the evening": if the shift ends at five and the manager asked someone to stay until seven, there are two overtime hours; if the employee stays on their own initiative to get ahead, there is no overtime, even if the result benefits the company.
The distinction matters because the rules are different. A premium is paid for working at an unfavorable moment —at night, on a Sunday, on a holiday— even when the shift itself is normal. Overtime is paid because the shift ran long. The same shift can carry both: someone who works nights on their ordinary schedule earns a night premium; if they also stay two extra hours, those two hours are night overtime.
Travel time, breaks the worker takes at will and on-call availability without effective work sit in a grey area: they depend on the role, the agreement and the internal rules, and it is better to define them before a conflict than after.
The four premium rates you have to tell apart
The whole calculation comes down to identifying which of these five rows each hour falls into. The percentages are the legal ones; the factor is simply the percentage turned into a multiplier so the arithmetic can be done.
| Type of hour | Premium | Factor | When it applies |
|---|---|---|---|
| Daytime overtime | 25 % | 1.25 | Exceeds the shift and falls in daytime hours |
| Night overtime | 75 % | 1.75 | Exceeds the shift and falls in night hours |
| Sunday or holiday, daytime | 100 % | 2.00 | Ordinary shift worked on a Sunday or holiday |
| Sunday or holiday, night | 150 % in practice | 2.50 | Night shift that falls on a Sunday or holiday |
| Night premium without overtime | 35 % | 1.35 | Ordinary shift inside the night window |
The first three rows are not interchangeable: night overtime already includes the night condition and Sunday work already includes the day. Adding two rows that are already complete is where most of the differences found in a payroll review come from.
How to calculate the ordinary hourly rate
The whole calculation hangs on a single number: the value of the ordinary hour. There are two ways to reach it and both must give the same result.
By day: monthly salary ÷ 30 = value of the day; value of the day ÷ hours in the daily shift = value of the hour. With a salary of $2,400,000 and an 8-hour shift: $2,400,000 ÷ 30 = $80,000 per day, and $80,000 ÷ 8 = $10,000 per hour.
By month: monthly salary ÷ hours in the current monthly shift. If the shift is 8 hours a day, the monthly figure is 30 × 8 = 240 hours; $2,400,000 ÷ 240 = $10,000. The same number by the other route.
The classic mistake is mixing the two routes: dividing by 240 when the shift has changed, or dividing by 8 without first dividing by 30. It is also wrong to divide by the days actually worked in the month and then apply the factor to that base: the result comes out higher than the legal one, and the difference becomes a liability nobody provided for.
One detail is usually overlooked. The legal maximum working week is being reduced in stages down to 42 hours, so your monthly divisor changes even when the salary does not: the same person earns the same and their hour is worth more. Confirm the number of hours in the current year's shift before you run payroll, and use that shift's divisor, not last year's.
Daytime overtime: the 25 % and its most common trap
Daytime overtime carries a 25 % premium over the ordinary hour, factor 1.25. Five daytime overtime hours on a $10,000 hour are $10,000 × 1.25 × 5 = $62,500.
The trap is not in the percentage but in the classification. An hour worked at night is not daytime just because the shift started in daylight, and an hour paid as daytime at eight in the evening is a difference the employee can claim years later. The line between day and night is set by law at a specific time window that has been debated and adjusted in recent years: confirm the current boundary before classifying, especially with rotating shifts or when the employee clocks in before dark.
The second mistake is applying the 25 % to an hour that already carries a premium. The percentage is calculated on the clean ordinary hour. If the night condition is also present, the right concept is night overtime, not the sum of the two premiums.
Night overtime: the 75 % and why it is not added to the 35 %
Night overtime carries a 75 % premium, factor 1.75. Three night overtime hours on a base of $10,000: $10,000 × 1.75 × 3 = $52,500.
This is where the most expensive confusion appears. The 75 % already includes the fact that the hour is at night: you do not add the 35 % night premium on top. Paying 1.75 + 0.35 = 2.10 is overpaying; paying 1.35 is underpaying and leaving the difference as a debt. The rule is simple: when the hour is overtime, the overtime factor governs; the 35 % applies only to ordinary hours that fall inside the night window.
It is the mistake that repeats most in security, continuous production, restaurants and transport, where night is the normal shift and not the exception: there the entire payroll is touched by the night premium, not just the closing hours.
Sundays and holidays: the 100 % and the combination that gets forgotten
Daytime Sunday or holiday work carries a 100 % premium, factor 2.00: the hour is worth double the ordinary one. Eight daytime Sunday hours on $10,000 are $10,000 × 2.00 × 8 = $160,000.
The case that gets calculated badly is supplementary work. If, on top of working Sunday, the employee stays additional hours, those hours carry the Sunday premium plus the overtime premium: daytime overtime on a Sunday lands at 2.25 (100 % + 25 %). "I worked Sunday" is not the same as "I worked Sunday and stayed two hours longer", and the difference shows up in the detail of the settlement.
The combination of night work on a Sunday or holiday is the most debated one. In practice it is settled with a total premium of 150 % over the ordinary hour —factor 2.50— when the hour is at night and falls on a Sunday or holiday. If those same hours also exceed the shift, the factor rises above 2.50, because Sunday and night are then compounded by the overtime condition. The components are not always added by simply summing each percentage: this is the point where you should review the table your system applies and confirm it with your accountant before paying.
Night premium without overtime: the 35 % almost nobody pays
When an employee works at night inside their normal shift there is no overtime: there is a 35 % night premium, factor 1.35. It is the concept most often omitted because it does not create a visible line in the shift; it hides inside the normal payment, which is why nobody reviews it.
It affects companies with rotating shifts, shifts that start before dawn, or staff who close at midnight. If your payroll pays the same to someone working 8 to 5 as to someone working 10 pm to 6 am, there is an accumulated liability, and it is not a rounding detail: it is several percentage points on every night hour, month after month.
The night premium is also salary. It enters the base for social contributions and benefits just like overtime, even when it does not appear as a separate line on the payslip.
The limits: 2 hours a day and 12 a week
Overtime has a ceiling: 2 hours a day and 12 a week. Beyond that the work cannot be required, and paying for it does not validate the excess. It is worth remembering when operations ask for "one more week" and the record shows four overtime hours a day for a whole month.
The limit is not paperwork. Beyond the sanction risk, sustained excess is the first piece of evidence to appear in a case: the payroll shows the pattern and the company cannot prove the work was voluntary.
The other temptation is offsetting overtime with time off. That only works with an express, verifiable agreement between the parties, and even then it has rules that depend on the specific case. Paying with "a day off next week" on a verbal promise is the fastest way to owe the hours twice.
Written authorization: no record, no proof
All overtime must be authorized in writing and recorded. This is not bureaucracy: it is the only way the company can show it knew about and approved the supplementary work, and the only way the employee can collect what is owed without an argument.
The authorization can be an email, a work order, a signed form or the system log, as long as it is clear who gave it, for which date and with what limit. A verbal instruction from the immediate manager also binds the company if it can be proven, but without written support the case becomes a contest of testimonies.
The record must allow the week to be reconstructed: clock-in, clock-out, breaks and who approved. If the system only keeps monthly totals, any claim turns into an argument about figures nobody can verify.
A full example with all four types of hour
Take a monthly salary of $2,400,000 and an 8-hour daily shift, with the ordinary hour of $10,000 calculated above. So that the arithmetic reads without trailing zeros, the exercise works with an hour of $100,000 —ten times that base—: the multipliers are the same, and if your hour is worth $10,000 you simply divide each total by ten.
| Type of hour | Quantity | Factor | Calculation | Total |
|---|---|---|---|---|
| Daytime overtime | 5 | 1.25 | $100,000 × 1.25 × 5 | $625,000 |
| Night overtime | 3 | 1.75 | $100,000 × 1.75 × 3 | $525,000 |
| Sunday daytime (shift) | 8 | 2.00 | $100,000 × 2.00 × 8 | $1,600,000 |
| Sunday night | 2 | 2.50 | $100,000 × 2.50 × 2 | $500,000 |
| Total for the period | 18 | — | — | $3,250,000 |
At the real rate of $10,000, that same period pays $325,000. The structure does not change: only the base changes.
Two warnings about this table. First: the percentages are legal, the amounts are not. If your company pays above the legal minimum, the table still works, but the result will be higher. Second: the eight Sunday hours in row three are an ordinary shift, not overtime; that is why they use 2.00 and not 2.25. If those hours had exceeded the shift, the factor would be different.
How to record hours so the calculation holds up
The settlement is the easy part. The hard part is having the support when someone asks why that week paid more than the previous one. A useful record answers five questions and nothing more.
| What to record | Why |
|---|---|
| Exact clock-in and clock-out, per day | Lets the week be reconstructed and the 2-hour daily excess spotted |
| Type of hour settled (daytime, night, Sunday) | Prevents the classification being done from memory at month end |
| Who authorized it and when | Without this, the company has nothing to oppose a claim |
| Breaks and on-call time | The two areas where the total is most disputed |
| Base salary and divisor used | If the divisor changes, the hourly rate changes even when the salary does not |
A weekly close is better than a monthly one. A week reviewed on time is fixed with a conversation; a month closed with errors is fixed with a payroll adjustment, and a year closed with errors is fixed in court.
The effect on contributions, service bonus, severance and vacation
Overtime is salary. It is not an allowance or a bonus: it enters the base for health, pension and workplace risk contributions, and also for the service bonus, severance and vacation. For vacation it is settled as an average, not as the value of the last month.
That is why "saving" on overtime is an illusion. The difference not paid today is still a debt tomorrow, with two aggravating factors: it is recalculated on bases that also grew and it carries interest. A cut of $200,000 a month for two years is not a saving of $4,800,000; it is a liability of that amount plus whatever accumulates in benefits.
The night premium follows the same path. If accounting and HR keep separate books, that is exactly where both lose track of it.
Mistakes that end up in court
Payroll mistakes are not litigated over the amount, they are litigated over the pattern. These are the ones that keep appearing, with the symptom that gives them away.
| Mistake | Symptom | Consequence |
|---|---|---|
| Wrong base | The hour is worth different amounts depending on who runs payroll | Retroactive differences for every month affected |
| Forgetting the 35 % night premium | Night shift and day shift pay the same | A claim covering every night hour |
| Paying Sunday as an ordinary day | The Sunday payroll looks "normal" | The full 100 % premium, plus benefits |
| Adding 75 % + 35 % on night overtime | Totals do not match the factor | Overpayment that cannot be recovered later |
| No authorization and no record | The employee supplies their own count | The company loses the argument for lack of evidence |
| Offsetting with time off, no agreement | Days off nobody documented | The hours are owed again, in full |
None of this requires bad faith: most of it comes from a spreadsheet copied from one year to the next and a shift change nobody updated in the divisor.
An honest closing before you decide
This article explains the mechanics, it does not resolve your case. Collective agreements, shift pacts, internal rules and sector situations change the result, and there are open debates —the night window, how premiums are added, holiday work— where your company's practice and your accountant's judgment weigh more than a general rule.
Before changing the settlement, confirm the number of hours in the current year's shift, check whether the divisor your payroll uses matches that shift, and verify that the time record allows a full week to be reconstructed. If the result of this exercise differs from what you pay today, the difference is the size of the problem.
When time records and payroll cost live in the same system, as in Kardex Tauro, that cross-check stops depending on a loose spreadsheet and can be reviewed before the month closes. If your case involves an ongoing proceeding or a liability of several years, this does not replace your lawyer or your accountant: they are the ones who have to look at the documents.
