Kingsford Pumps: cross-subsidy, a special order and a costing method under pressure
Kingsford Pumps manufactures two lines at one plant. The Standard line sells 30,000 units a year and uses 2 direct labour hours each; the Custom line sells 2,000 units a year and uses 10 direct labour hours each. Total manufacturing overhead is $4,800,000, applied on direct labour hours (80,000 hours, a plant-wide rate of $60 per hour). Standard's direct materials are $180 and direct labour $44 per unit; Custom's are $700 and $220. Pricing is cost-plus 25% on full manufacturing cost. A costing review produced four activity pools: setup $960,000 (800 setups: 200 Standard, 600 Custom); machining $2,400,000 (40,000 machine hours: 30,000 Standard, 10,000 Custom); materials handling $720,000 (1,800 moves: 600 Standard, 1,200 Custom); inspection $720,000 (1,200 inspections: 300 Standard, 900 Custom). The Custom sales manager has just won a three-year order for 400 units a year at $1,850, describing it as 'profit at last'. The sales director wants overhead reallocated on machine hours because 'direct labour is shrinking'.
1. Recalculate unit costs using activity-based costing, quantify the cross-subsidy, and identify two decisions the plant-wide rate would distort.
Read the worked answer
Activity rates: setup $960,000 / 800 = $1,200 per setup; machining $2,400,000 / 40,000 = $60 per machine hour; handling $720,000 / 1,800 = $400 per move; inspection $720,000 / 1,200 = $600 per inspection.
Overhead absorbed - Standard: 200 x 1,200 = 240,000; 30,000 x 60 = 1,800,000; 600 x 400 = 240,000; 300 x 600 = 180,000. Total 2,460,000, or $82 per unit (30,000 units).
Overhead absorbed - Custom: 600 x 1,200 = 720,000; 10,000 x 60 = 600,000; 1,200 x 400 = 480,000; 900 x 600 = 540,000. Total 2,340,000, or $1,170 per unit (2,000 units).
Cross-subsidy: the plant-wide rate charges Standard $120 and Custom $600. Standard is over-costed by $38 x 30,000 = $1,140,000; Custom is under-costed by $570 x 2,000 = $1,140,000. The two offset exactly, which is why total plant profit looked fine.
Full cost and current price: Standard $180 + $44 + $82 = $306 (plant-wide $344), cost-plus price $382.50 (plant-wide $430). Custom $700 + $220 + $1,170 = $2,090 (plant-wide $1,520), cost-plus price $2,612.50 (plant-wide $1,900). On ABC, Custom's current $1,900 price is about 9% below full cost.
Decisions distorted:
- Pricing and order acceptance. Standard has been priced at $430 against competitors near $400, losing volume, while Custom has been won at prices below cost. Every Custom order at $1,900 destroys value.
- Product mix and capital allocation. The plant-wide rate shows Custom at a 20% gross margin, so investment, capacity and sales incentives flow to the line that is actually loss-making, and any make-or-buy or line-rationalisation analysis inherits the same error.
2. Using relevant costing, advise whether to accept the $1,850 order for 400 units a year for three years. Quantify and state your assumptions.
Read the worked answer
First test capacity. Current activity usage against capacity: setups 600 of 800 (75%); inspections 900 of 1,200 (75%); moves 1,200 of 1,800 (67%); machine hours 10,000 of 40,000 (25%). The order adds 120 setups (400 x 0.3), 180 inspections, 240 moves and 2,000 machine hours, taking setups and inspections to 90% utilisation. Capacity exists, but the buffer is thin.
Relevant cost per unit. If batch resources are genuinely spare, the incremental cost is direct materials $700 + direct labour $220 = $920. Contribution = $1,850 - $920 = $930 per unit, or $372,000 a year, $1,116,000 over three years undiscounted.
If any batch resource must be added - for example a setup technician at $95,000 and 0.2 of an inspector at $30,000 - annual contribution falls to about $247,000 but stays positive. Even at 90% utilisation the order still clears its own costs.
Why ABC full cost of $2,090 does not govern. Full ABC cost is the right floor for sustained, normal-volume pricing because it recovers the fixed activity capacity the business must hold. It is not the relevant cost of a defined-term order that consumes otherwise idle capacity. The trap is to treat a $240 headline loss as a reason to refuse.
Recommendation: accept, subject to conditions. (1) Confirm in writing that no step cost is triggered; if it is, add it to the price analysis. (2) Include a 12-month price review and an indexation clause - three years at a fixed price against rising wages is the real risk. (3) Ring-fence the price: no most-favoured-customer clause, and use a distinct specification or private label so the $1,850 does not become the reference price for the 2,000 units sold at $1,900 and above. (4) Confirm no displacement of existing Custom orders. (5) Disclose to the board that the base Custom business is loss-making on ABC and needs repricing regardless of this order.
3. Evaluate the sales director's proposal to allocate overhead on machine hours instead, and recommend how Kingsford should set its product prices.
Read the worked answer
The proposal trades one distortion for another. A machine-hour rate of $4,800,000 / 40,000 = $120 per hour would charge Standard 1 machine hour x $120 and Custom 5 hours x $120 = $600. That simply reproduces the plant-wide result for Custom and is no better justified. Machine hours are a driver of the machining pool only; they are not a driver of setup, handling or inspection work, which is precisely where Custom consumes most of its overhead. The stated reason - that direct labour is shrinking - is an argument against using a labour-based driver, not an argument for machine hours.
Method choice should follow decision use, cost and behaviour. Keep the plant-wide rate out of pricing entirely. Use ABC, reviewed annually, for pricing, product mix, quoting and profitability by line and by customer. For monthly management reporting, a two-driver model (machining by machine hours, batch activity by setups) captures most of the distortion at a fraction of the administrative cost. Never change a costing method because it produces a more comfortable number; that is a decision-support failure, and if it were done to support a predetermined price it would also raise an objectivity issue.
Pricing recommendation.
- Standard: priced at $430 against a competitor at $400. Cut to around $380-$390. ABC full cost is $306, so contribution remains strong and volume recovery should outweigh the margin give-up. Test elasticity with a regional price trial before a national cut.
- Custom: raise prices 10-15% toward the $2,300-$2,400 range and accept that some volume will go. At $2,090 full cost, a 10% rise still leaves a thin margin; the objective is to stop subsidising the line.
- Segmentation: where possible, offer a simplified Custom variant with standard sub-assemblies, cutting setups and inspections per unit - attacking the cost driver is better than arguing about the allocation.
- Governance: require quotes below ABC full cost to be approved by the CFO, with a documented strategic justification. This single control would have caught the $1,850 order earlier and will catch the next one.