When every unit is a clone: why job-order costing breaks down
In the previous guide of this rung you met job-order costing, which works beautifully when each unit is *distinguishable* — a custom yacht, a single architect's project, a batch of printed wedding invitations. There you open a job cost sheet, heap direct materials, direct labour, and overhead onto that one job, and when it is done you divide by the units in the job to get a cost each. The whole method rests on being able to point at a job and say *that one*. Now ask what happens at a paint factory, a flour mill, or a chemical plant.
In those places production is one continuous river. Litre after identical litre of white paint pours down the line; one kilogram of flour is indistinguishable from the next; the chemical reactor never makes a *batch* so much as a *flow*. There is no "job number 47" to point at — there is only May, during which the plant consumed a heap of materials and labour and produced, say, 100,000 litres. Trying to track the cost of each individual litre would be absurd and pointless: the litres are clones, so the cost of any one is, by definition, the average cost of all of them. This is the world of [[process-costing|process costing]] — costing by *department over a period*, not by *job*.
The trouble with half-finished work, and the trick that solves it
Averaging sounds easy — total cost divided by units made. But here is the snag that makes process costing genuinely subtle. At the end of May, the paint line does not stop at a tidy boundary. Some paint is fully finished and shipped out; but a lot of it is still in the pipes and mixing tanks, *half-made*. If the department spent enough money to make 100,000 litres' worth of effort, but only 80,000 litres walked out the door fully done while 40,000 litres sit half-finished, what exactly do you divide your costs by? Dividing by the 80,000 finished alone overstates the cost per litre; dividing by all 120,000 physical litres pretends the half-done ones cost as much as the finished — also wrong.
The resolution is one of the most elegant ideas in all of cost accounting: the [[equivalent-units|equivalent unit]]. Instead of counting physical units, we count *units of effort*. The rule is plain arithmetic: ten units that are each 50% complete represent exactly five whole units' worth of work — 10 × 50% = 5 equivalent units. Forty thousand litres that are 25% done embody the same effort as 10,000 finished litres. We are not lying about how many physical litres exist; we are translating partly-done work into the equivalent number of fully-done units, so that an honest average — a cost *per unit of work actually performed* — can finally be computed.
The weighted-average method, worked end to end
The most common way to put equivalent units to work is the [[weighted-average-process-method|weighted-average method]]. Its spirit is exactly the weighted-average cost idea you already met for inventory: it refuses to keep this month's costs separate from last month's leftovers. It simply pools the cost that was already sitting in beginning work-in-process together with all the cost added this period, then spreads that single blended pool across all the work done — measured in equivalent units. Old effort and new effort, old money and new money, all averaged into one smooth cost per equivalent unit.
Let us make it concrete with one mixing department in May. It finished and transferred out 80,000 litres, and left 40,000 litres in ending work-in-process that are 100% complete for materials but only 25% complete for conversion. The cost per equivalent unit comes in two streams. For materials, the equivalent units are 80,000 (finished) + 40,000 (the leftovers, fully materialed) = 120,000; if total material cost in the pool is 240,000, the cost is 2.00 per litre. For conversion, the leftovers count only 40,000 × 25% = 10,000 equivalent units, so conversion equivalent units are 80,000 + 10,000 = 90,000; if the conversion-cost pool is 270,000, the cost is 3.00 per litre. A fully finished litre therefore costs 2.00 + 3.00 = 5.00.
WEIGHTED-AVERAGE, Mixing Dept, May Materials Conversion
Equivalent units of work done
Finished & transferred out 80,000 80,000 80,000
Ending WIP 40,000 units
materials 100% 40,000
conversion 25% 10,000
------------------------------------ ---------- ----------
Equivalent units 120,000 90,000
Cost to account for (pool) 240,000 270,000
Cost per equivalent unit = 240,000/120,000 = 2.00
270,000/ 90,000 = 3.00
Cost of one finished litre = 2.00 + 3.00 = 5.00
Where the pool goes:
Transferred out 80,000 x 5.00 = 400,000
Ending WIP materials 40,000 x 2.00 = 80,000
conversion 10,000 x 3.00 = 30,000 = 110,000
--------------------------------------------------------
Total accounted for = 510,000 (= 240,000 + 270,000)How the cost flows: WIP, finished goods, and the next department
Now watch the money actually move, because process costing is ultimately a story of cost flowing through accounts. Each department keeps its own Work-in-Process account. Costs pour *in* — direct materials requisitioned, labour, and applied overhead. The cost per equivalent unit then governs how much pours *out*: the finished litres carry their cost out of this department's WIP, while the partly-done litres leave their cost behind as the ending WIP balance, which becomes next month's beginning balance. Nothing is lost; the WIP account is simply a tank with cost flowing in at the top and out the bottom.
Here is the feature unique to process costing: most products pass through *several* departments in sequence — mixing, then tinting, then canning — and the cost transferred out of one department flows in as a starting cost of the next, where it is called transferred-in cost. By the time the paint reaches the last department, its WIP carries the accumulated cost of every step before it. When that final department finishes a litre, its full cost moves out of WIP and into the Finished Goods account; when the litre is finally sold, that same cost moves once more, from Finished Goods into cost of goods sold on the income statement. The cost has now completed its journey from raw material to expense.
- Summarize the physical flow: units in beginning WIP plus units started must equal units finished-and-transferred plus units in ending WIP.
- Compute equivalent units separately for materials and for conversion, using each leftover's own completion percentage.
- Compute the cost per equivalent unit: take the cost pool for each and divide by its equivalent units; add the two to get the cost of one finished unit.
- Assign the cost: finished units take the full cost out to the next department or finished goods; ending WIP keeps cost equal to its equivalent units times the per-unit costs — and check that out plus in equals the total to account for.
The FIFO method, and honest caveats
Weighted-average has a known blind spot: by blending last month's leftover costs with this month's, it blurs the two periods together, so it can hide whether *this* month's work actually got cheaper or dearer. The [[fifo-process-method|FIFO process method]] fixes that by keeping the periods clean and separate. Just like the FIFO cost-flow assumption you met for inventory, it insists the beginning work-in-process is finished *first*, using its own already-incurred cost, before any newly started unit is touched. Its equivalent-unit count therefore includes *only the work done this period*: the effort to finish off the beginning WIP, plus the units started and finished, plus the work done so far on the ending WIP.
The practical difference is usually small and the two methods agree when input prices are stable or when beginning WIP is tiny. Weighted-average wins on simplicity and is far more common in practice; FIFO wins on precision for cost control, because its per-unit cost reflects *only* current-period performance, which is exactly what a manager wants when judging whether this month's process ran efficiently. Many real systems quietly use weighted-average and reserve FIFO for when a clean current-period figure genuinely matters. Neither is "more correct" — they answer slightly different questions about the same river of product.
Three honest caveats before you leave. First, the completion percentages are *estimates* — an engineer's judgement that the tanks are "about 25% cooked," not a measured fact — so a process cost per unit is only as trustworthy as those guesses; deliberately nudging them is a classic way to massage reported profit. Second, the cost per equivalent unit is an *average*, and like every average it conceals variation: it cannot tell you that the night shift wasted material while the day shift did not. Third, process costing tells you what a unit *cost on average*, never what it *should* have cost — that comparison waits for the standard costs and variance analysis of the next rung, which finally asks whether 5.00 a litre was good, bad, or alarming.