When a factory runs, costs like rent, electricity, and the canteen don't belong to just one department — they belong to everyone. Apportionment is the process of sharing these common overhead costs across departments using a fair basis. The goal is simple: every product should carry its fair share of the factory's total costs.
In a typical factory, you have two kinds of departments. Production departments (like Machining, Assembly, Finishing) directly work on products. Service departments (like Stores, Maintenance, Canteen, Power House) support production but don't touch the product themselves. In primary apportionment, you spread all overheads — both production and service department costs — across all departments using appropriate bases of apportionment: floor area for rent, machine hours for power, number of employees for canteen, and so on. After primary apportionment, your service departments have a pile of cost sitting with them — but service departments don't make products, so those costs must eventually reach the production departments. That's secondary apportionment (also called re-apportionment).
For re-apportionment, ICAI tests three main methods. The Direct Method is the simplest — ignore any services one service department gives to another, and push all service dept costs straight to production departments only. The Step (Sequential) Method closes service departments one by one in a sequence; a closed department doesn't receive any further cost. The Repeated Distribution Method keeps cycling costs between service departments until the numbers become negligible (usually till the remainder is ₹1 or zero) — this is the most accurate when service departments serve each other significantly. The Simultaneous Equation Method sets up algebraic equations to solve for the true total cost of each service department, accounting for mutual services mathematically. This is asked frequently as a 6–8 mark question in Paper 4, and the examiner loves the Repeated Distribution and Simultaneous Equation methods for their complexity.