Two Rates in One Week
Computing a premium for somebody who worked at more than one rate, why systems default to the wrong answer, and who is multi-rate without knowing.
One week at two rates, as one employer's system computed it
The system used whichever rate the employee happened to be on when they crossed the threshold. How multi-rate premiums must be computed differs by jurisdiction and is a question for somebody qualified in the place concerned.
Where somebody works at more than one rate in a week, the premium is usually computed on a weighted average of the hours actually worked at each — not on whichever rate they happened to be on when they crossed the threshold. That default is arbitrary, it is what several systems do, and it produces a different answer depending on the order of the shifts.
The pay calculation in “Two Rates in One Week” depends on a complete record before any rate is applied. For organisations researching how to calculate idle time, this workforce tool can connect hours with projects and review steps, provided payroll keeps the governing formula, contract terms and disputed-entry process outside any single dashboard.
The calculation itself is simple. The problem is that almost nobody knows they have multi-rate employees, because the second rate was introduced for a different reason and nobody connected it to the premium.
For an independent reference relevant to “Two Rates in One Week”, consult the SAM.gov federal award resources. Use it to test record quality, working-time definitions, access, retention and exception handling against the organisation’s real payroll process.
The arithmetic
- Total the pay for all hours worked in the week, at each rate.
- Add any other components that belong in the rate.
- Total the hours actually worked.
- Divide the pay by the hours: this is the weighted average rate.
- Apply the premium multiplier to that rate for the premium hours.
- Check what the system used instead.
Step six is the one that matters. In a surprising number of systems the answer is the rate on the last entry, the rate on the highest-paid entry, or the employee's "primary" rate — none of which is the weighted average and all of which are order-dependent.
Who is multi-rate without knowing
- Anybody who covers a second role, even occasionally.
- Anybody whose travel time is paid at a different rate from their work.
- Anybody paid a training rate while learning and a full rate afterwards, within the same week.
- Anybody who works at two sites with different site rates.
The second is the one that catches organisations out, because paying travel at a lower rate feels like a cost decision rather than a pay structure change. It is both.
The order dependency, demonstrated
Two shifts, the same hours, the same rates, different order: in a system that uses the last entry, the premium is computed on the lower rate if the cheap shift came last and the higher rate if it came first.
An employee cannot be paid differently for the same week depending on which day the rota put the training shift. That is enough on its own to show the configuration is wrong, without any reference to a rule.
Finding them
Query the payroll data for anybody with more than one rate code in the same week. Most systems can answer this and nobody has ever asked.
The list is usually longer than expected and clusters: a department that runs a training rate, a service that pays travel separately, a site with a shortage supplement. Each cluster is one configuration decision rather than a set of individual problems.
Where an alternative method applies
Some systems permit an agreed alternative — paying the premium at the rate applicable to the work being done when the extra hours occurred, where that has been agreed in advance.
Whether that is available, and what the agreement has to look like, differs by jurisdiction and is a question for somebody qualified in the place concerned. What is not available anywhere is applying an alternative method by accident because the system defaulted to it.
Testing the configuration
Build one artificial week: twenty hours at one rate, twenty-two at another, run it through the system and compare with the hand calculation. Then reverse the order of the shifts and run it again.
Three numbers, ten minutes, and it answers the question for every multi-rate employee in the organisation. It is also the test to rerun after any payroll upgrade, because this is exactly the kind of behaviour that changes quietly in a release.
Two employers in one group
Somebody who works for two companies in the same group, in the same week, raises the multi-rate question with an extra element: whether the hours combine at all for the threshold.
That depends on the relationship between the entities and is a question for somebody qualified in the place concerned. The operational step is to find out whether it happens, which requires comparing payroll across entities — something most groups have never done and which frequently turns up more people than anybody expected.
Acting on the test result
If the system is right, record the test and the date. If it is wrong, the difference applies to every multi-rate premium hour it has ever computed, which is a quantity somebody needs to establish before deciding what to do about it.
Either way the fix is the same shape: correct the configuration, then work out how far back it reaches. The second half is harder and is the subject of its own section here.