Make-or-buy decisions

Why it matters

Almost every manufacturer has to decide, part by part, whether to produce in its own shop or buy from a vendor: castings, fasteners, PCBs, gears, heat-treatment, even whole sub-assemblies. A wrong decision either ties up capacity and capital in work a specialist does better, or hands a critical capability to a supplier. The economics are simple once you separate relevant from irrelevant costs, but the decision also has strategic and quality sides.

Key ideas

Relevant costs. Only costs that differ between the alternatives matter. For "make" these are usually the variable costs (direct material, direct labour, variable overhead) plus any fixed costs that are incurred only if the part is made — a new fixture, die or machine, an extra supervisor. Fixed overheads that the firm pays anyway (factory rent, existing depreciation, general administration) are sunk or unavoidable and must be left out, even if the cost sheet allocates them to the part.

Relevant cost of buying. Purchase price plus freight, incoming inspection, ordering and any duties, minus any fixed costs that would be saved by stopping in-house production.

Capacity and opportunity cost.

• With idle capacity, making uses resources that would otherwise earn nothing, so the opportunity cost is zero.
• With full capacity, making the part displaces another product. The contribution lost from that product is an opportunity cost that must be added to the cost of making. It is usually computed per unit of the scarce resource (e.g. contribution per machine-hour × hours per part).

Break-even quantity between make and buy. If making requires an avoidable fixed cost F_m and variable cost v_m, and buying costs p per unit, the two total-cost lines cross at Q_e = F_m / (p − v_m) (only meaningful if p > v_m). Below Q_e buy; above Q_e make. This is the same cost-indifference idea as in break-even analysis.

Qualitative factors.

• Favouring make: protecting proprietary design or process know-how, tight quality or delivery control, core competence, using idle capacity and keeping a skilled workforce, unreliable or monopoly suppliers.
• Favouring buy: small or fluctuating volumes, the supplier's specialist expertise and economies of scale, lack of capacity or capital, faster access to new technology, freeing management attention for core products.
• Risks of buying: supply disruptions, price increases once dependent, IP leakage, longer lead times. Risks of making: under-used investment if volumes fall, being locked into an old technology.

Decisions change over a product's life: a firm may buy during launch while volumes are uncertain and make once volume is stable (or the reverse as a part becomes a commodity).

Formulas

• Relevant cost to make (per unit): c_make = v_DM + v_DL + v_VOH + opportunity cost per unit
• Total make cost: TC_make = F_m + c_make·Q
• Total buy cost: TC_buy = (p + c_freight + c_inspection)·Q
• Make–buy indifference quantity: Q_e = F_m / (p − c_make)
• Opportunity cost per unit (scarce machine hours): c_opp = (contribution per hour of displaced product) × (hours per part)
• Saving from making: S = TC_buy − TC_make (make if S > 0)

Symbols: F_m = avoidable fixed cost of making (₹ per period); v_DM, v_DL, v_VOH = variable material, labour and overhead per unit (₹/unit); p = purchase price (₹/unit); Q = annual requirement (units); Q_e = indifference quantity (units).

Worked examples

Example 1 (standard) — ignore unavoidable fixed overhead. A firm buys a valve body at ₹120 each and needs 10,000 per year. Its cost sheet for making it shows: direct material ₹50, direct labour ₹30, variable overhead ₹15, fixed overhead absorbed ₹25 — total ₹120. The fixed overhead is existing factory overhead that will not change. The shop has idle capacity. Should it make or buy?

1. The ₹25 fixed overhead is incurred whether or not the part is made, so it is irrelevant.
2. Relevant make cost = 50 + 30 + 15 = ₹95 per unit.
3. Saving from making = (120 − 95) × 10,000 = ₹2,50,000 per year.

Decision: make. The full-cost comparison (₹120 vs ₹120) would wrongly suggest indifference.

Example 2 (GATE level) — new fixture and full capacity. Same part, but making it now requires a special fixture costing ₹1,50,000 per year (avoidable), and the machine is fully loaded. Each valve body takes 0.5 machine-hour, and the product it would displace earns a contribution of ₹40 per machine-hour.

1. Opportunity cost per unit = 40 × 0.5 = ₹20.
2. Relevant variable make cost = 95 + 20 = ₹115 per unit.
3. TC_make = 1,50,000 + 115 × 10,000 = ₹13,00,000. TC_buy = 120 × 10,000 = ₹12,00,000.
4. Decision for 10,000 units: buy, saving ₹1,00,000.
5. Indifference quantity = 1,50,000 / (120 − 115) = 30,000 units per year; making would pay only above this volume.

(Had the machine been idle, the indifference quantity would be 1,50,000 / (120 − 95) = 6,000 units, and making 10,000 would cost ₹11,00,000 — cheaper than buying.)

Common mistakes

• Comparing the purchase price with the full absorbed cost, including fixed overheads that will not disappear.
• Ignoring the opportunity cost when capacity is fully used.
• Leaving out buying costs other than price: freight, inspection, rejections, inventory carrying.
• Treating Q_e = F_m / (p − c_make) as valid when p < c_make (then buying is always cheaper).
• Deciding on cost alone for a part that embodies core know-how or has a single unreliable supplier.

For GATE PI

Typical questions give a make cost structure and a purchase price and ask for the decision, the saving, or the indifference quantity — often with a trap fixed-overhead figure that must be excluded. Conceptual MCQs ask which factors favour make or buy. Practise identifying avoidable versus unavoidable costs and adding opportunity cost when capacity is tight.

Quick check

1. Make: avoidable fixed ₹40,000, variable ₹180 per unit; buy: ₹220 per unit. Find the indifference quantity.
2. Is allocated factory rent relevant to a make-or-buy decision?
3. Name two qualitative reasons to keep a part in-house.
4. When does opportunity cost enter the make cost?

Answers: 1. 40,000 / 40 = 1,000 units; 2. no, it is unavoidable; 3. protecting proprietary know-how and controlling quality or delivery (also core competence, idle capacity); 4. when making the part displaces other profitable work because capacity is fully used.