Three loss-of-pay days on ₹26,000 gross. What is the deduction?
With a 26-day divisor it is ₹3,000. With 30 it is ₹2,600. With the actual days in a 31-day month it is ₹2,516.
All three are defensible. What is not defensible is using different ones in different months.
Consistency is the whole issue
Employees compare payslips — with each other, and with their own from three months ago. A per-day rate that moves without explanation reads as arbitrary, and once payroll is seen as arbitrary, every other number in it is questioned too.
That cost is much larger than the few hundred rupees in dispute, and much harder to recover.
Choosing one
26 suits establishments working a six-day week, where it approximates working days excluding weekly offs. It is common in manufacturing.
30 is simple, predictable, and produces the same rate every month regardless of length. Employees find it easy to check, which is worth something.
Actual days is arithmetically precise and produces a rate that varies between February and March — which is correct and generates questions anyway.
There is no universally right answer. There is only a stated answer applied to everyone.
Write down the other half too
The divisor is one decision. The base is the other: is LOP computed on gross, or on basic? That materially changes the deduction, and it should be in the policy alongside the divisor.
And state what LOP does to other things — whether it affects leave accrual, and how it appears in the days-worked figure reported in your PF ECR, which must agree with the register.
What LOP touches downstream
It reduces wages for the month, and therefore the PF and ESI computed on them. It reduces the days worked reported in the ECR. Both need to follow automatically from the same LOP figure — if the deduction is applied in payroll but the days reported in the return come from somewhere else, they will eventually disagree.
The check
Take one employee with LOP in three different months and compute the implied per-day rate for each. If the three do not match, you have found something worth fixing before somebody else does.