Production and costs

In short

Production theory explains how a firm converts inputs (land, labour, capital, enterprise) into output, and cost theory converts that physical relationship into rupees. Together they decide how much a firm should produce, what plant size to build, which input mix is cheapest, and the price below which production must stop. This is the core of Microeconomics papers under UGC CBCS, of Managerial Economics in MBA entrance and of UPSC/RBI/NABARD/UGC-NET economics sections. Analysts at consulting firms, FMCG and manufacturing companies, banks and regulators (CCI, CERC, TRAI) use exactly these cost concepts for pricing, capacity and tariff decisions.

Explained

Production function and time periods

• Production function: the technical relation giving maximum output from given inputs with given technology — Q = f(L, K, N, E). Technology is held constant; a change in technology shifts the whole function.
• Short run: at least one input (usually capital/plant) is fixed; output changes only by varying labour and materials. Long run: all inputs variable, no fixed cost. Very long run: technology itself changes.
• Fixed vs variable factors: plant, machinery, managerial staff are fixed in the short run; raw material, power, casual labour are variable.
• Total, average, marginal product: TP = Q; AP = TP/L; MP = ΔTP/ΔL (or dTP/dL).
• Relations: when MP > AP, AP rises; when MP < AP, AP falls; MP cuts AP at AP's maximum. TP is maximum where MP = 0, and falls when MP < 0.

Law of variable proportions (short-run returns to a factor)

• Statement: as more units of one variable factor are applied to a fixed factor, MP first rises, then falls, and finally becomes negative.
• Stage I — increasing returns: MP rises; better utilisation of the fixed factor and indivisibility of plant. AP rises.
• Stage II — diminishing returns: MP positive but falling, from AP's maximum to MP = 0. A rational firm always operates here (Stage II), because in Stage I the fixed factor is under-used and in Stage III output itself falls.
• Stage III — negative returns: too many workers crowd the fixed factor; MP < 0, TP falls.
• Causes: scarcity and indivisibility of the fixed factor, imperfect substitutability of factors.
• Assumptions: constant technology, homogeneous variable factor units, some factor fixed, output measurable.

Isoquants, factor substitution and least-cost combination

• Isoquant (iso-product curve): locus of input combinations giving the same output. Properties: negatively sloped, convex to the origin, never intersect, higher isoquant = higher output.
• MRTS(L,K) = −ΔK/ΔL = MPL/MPK — the rate at which capital can be given up for one more unit of labour, output constant. It diminishes along an isoquant, giving convexity.
• Special cases: perfect substitutes → straight-line isoquant; perfect complements (fixed proportions, Leontief) → L-shaped isoquant.
• Isocost line: C = wL + rK; slope = −w/r. Higher isocost = higher expenditure.
• Producer's equilibrium (least-cost combination): isocost tangent to isoquant → MRTS = w/r, i.e. MPL/w = MPK/r. Second-order condition: isoquant convex at the tangency.
• Expansion path (isocline/scale line): locus of tangency points as outlay rises — the firm's efficient growth path and the source of the long-run cost curve.
• Elasticity of substitution σ = % change in K/L ÷ % change in MRTS; σ = 1 for Cobb-Douglas, 0 for fixed proportions, ∞ for perfect substitutes. CES production function allows any constant σ.

Returns to scale and homogeneity

• Returns to scale (long run): all inputs changed in the same proportion. If inputs rise λ times and output rises more than λ → increasing returns to scale (IRS); exactly λ → constant (CRS); less than λ → decreasing (DRS).
• A production function is homogeneous of degree n if f(λL, λK) = λⁿ f(L, K): n > 1 IRS, n = 1 CRS (linear homogeneous), n < 1 DRS.
• Cobb-Douglas Q = A L^α K^β: degree = α + β; α and β are the output elasticities and, under competition and CRS, the income shares of labour and capital; σ = 1. Euler's theorem: with CRS, Q = L·MPL + K·MPK — the product is exactly exhausted by factor payments.
• Economies of scale (internal): technical (larger, more efficient plant; indivisibilities), managerial (specialisation of management), commercial (bulk buying and selling), financial (cheaper credit for large firms), risk-bearing (diversification), marketing.
• Diseconomies of scale: managerial span of control, coordination and communication failure, falling worker morale.
• External economies/diseconomies: arise from the growth of the whole industry — common infrastructure, skilled labour pools, ancillary units (Tiruppur knitwear, Surat textiles, Pune auto cluster) versus congestion and rising local input prices.
• Returns to scale ≠ law of variable proportions: the first changes all inputs (long run), the second only one (short run).

Cost concepts an exam asks you to distinguish

• Explicit (accounting) cost: actual payments to outsiders — wages, rent, material, interest paid.
• Implicit cost: value of owner's own factors — imputed salary, rent on own premises, interest on own capital.
• Economic cost = explicit + implicit. Normal profit is the minimum implicit return needed to keep the owner in business; it is a cost, not a profit, in economics.
• Accounting profit = revenue − explicit cost; economic profit = revenue − (explicit + implicit). Economic profit is smaller.
• Opportunity cost: value of the next-best forgone alternative — the basis of all economic cost.
• Sunk cost: already incurred and unrecoverable; irrelevant to future decisions (a classic interview trap).
• Fixed (supplementary/indirect) cost: unchanged with output in the short run — rent, insurance, depreciation on time basis, salaried staff.
• Variable (prime/direct) cost: varies with output — materials, power, wages of casual labour.
• Other pairs: private vs social cost (the gap is the externality, basis of pollution regulation and CPCB norms), historical vs replacement cost, incremental vs sunk cost, out-of-pocket vs book cost, escapable vs unavoidable cost, controllable vs uncontrollable cost, urgent vs postponable cost.

Short-run cost curves

• TC = TFC + TVC. TFC is a horizontal line; TVC starts at the origin and is S-shaped (rises slowly, then steeply) because of the law of variable proportions.
• AFC = TFC/Q falls continuously and is a rectangular hyperbola (AFC × Q = constant), approaching but never touching the axis.
• AVC = TVC/Q and AC = TC/Q are U-shaped; AC lies above AVC and the vertical gap (AFC) narrows as output rises; AVC reaches its minimum at a *lower* output than AC.
• MC = ΔTC/ΔQ = dTC/dQ = dTVC/dQ — MC depends only on variable cost, so fixed cost never affects MC.
• MC cuts both AVC and AC at their minimum points from below. When MC < AC, AC falls; when MC > AC, AC rises.
• Short-run cost and product curves are mirror images: AVC = w/APL and MC = w/MPL (constant wage w). MC is minimum where MP is maximum; AVC is minimum where AP is maximum.
• Modern (empirical) view: engineering and statistical studies find a flat-bottomed, saucer-shaped AVC over a wide range of output with MC constant in that range, rather than a sharp U.

Long-run cost, LAC and plant choice

• In the long run there is no fixed cost; the firm chooses the plant size. LAC is the envelope of short-run average cost curves — it touches each SAC but is not tangent at SAC's minimum except at the LAC minimum itself.
• LAC falls over the range of economies of scale, is flat at the minimum efficient scale (MES) and rises with diseconomies; LAC is flatter than SAC — hence "planning curve".
• LMC cuts LAC at LAC's minimum; at the optimum output SMC = LMC = SAC = LAC.
• Natural monopoly: if LAC falls over the entire relevant range (power transmission, gas pipelines, railways), one firm supplies most cheaply — the reason for tariff regulation by CERC/SERCs and PNGRB rather than competition.
• Learning curve (experience curve): unit cost falls by a fixed percentage each time cumulative output doubles — distinct from scale economies, which depend on the rate of output.
• Economies of scope: producing two products together costs less than separately (dairy co-operatives making milk, ghee and powder).
• Shut-down and break-even: a firm continues in the short run if P ≥ AVC (it is covering variable cost and contributing to fixed cost) and exits in the long run if P < LAC.

Key points

• Production function - maximum output obtainable from given inputs with given technology.
• Short run - at least one factor fixed; long run - all factors variable, no fixed cost.
• Law of variable proportions - MP of a variable factor rises, falls, then turns negative; the firm operates in Stage II.
• MP-AP rule - MP cuts AP at AP maximum; TP is maximum when MP = 0.
• Isoquant - equal-output curve, convex because MRTS diminishes; MRTS = MPL/MPK.
• Least-cost rule - MPL/w = MPK/r, i.e. MRTS = w/r, at the isoquant-isocost tangency.
• Returns to scale - all inputs scaled equally; increasing, constant or decreasing.
• Cobb-Douglas - Q = A L^α K^β; α + β decides returns to scale; α, β are output elasticities; σ = 1.
• Economies of scale - technical, managerial, commercial, financial, risk-bearing; diseconomies arise from managerial limits.
• Economic cost - explicit + implicit; normal profit is a cost, so economic profit < accounting profit.
• Opportunity cost - value of the best forgone alternative; sunk cost is irrelevant to decisions.
• AFC - rectangular hyperbola, falls continuously; AVC and AC - U-shaped; MC - unaffected by fixed cost.
• MC-AC rule - MC cuts AVC and AC at their minima; AVC minimum comes before AC minimum.
• Cost-product mirror - AVC = w/AP, MC = w/MP; MC minimum where MP is maximum.
• LAC - envelope of SACs, the planning curve; LMC cuts LAC at its minimum.
• Minimum efficient scale - smallest output at which LAC is at its lowest.
• Shut-down point - P = minimum AVC in the short run; exit when P < LAC in the long run.
• Break-even output - fixed cost ÷ contribution per unit, where contribution = price − AVC.

Formulas and examples

Example 1 — TP, AP, MP and the three stages. A workshop has one fixed machine; labour varies.

• MP rises to L = 3 → Stage I (increasing returns).
• AP is maximum (13.0) at L = 3–4, where MP = AP (13 at L = 4) → Stage II runs from L = 4 to L = 7.
• TP is maximum (64) at L = 7 where MP = 0; beyond that MP < 0 → Stage III. Operate between L = 4 and L = 7.

Example 2 — full short-run cost schedule. TFC = ₹60.

• AVC is minimum (₹21) at Q = 4; AC is minimum (₹34) at Q = 5 — AVC bottoms out first, as theory requires.
• MC is minimum (₹16) at Q = 3, then rises and passes through AVC's minimum region and AC's minimum (MC = 26 at Q = 5 against AC = 34, crossing between Q = 5 and 6 where AC turns up).
• At price ₹25, the firm covers AVC at Q = 4 (AVC 21) and keeps producing even though AC = ₹36 > ₹25: loss = (36 − 25) × 4 = ₹44, which is less than the ₹60 fixed cost it would lose by shutting down.

Example 3 — cost function by calculus. TC = 100 + 20Q − 4Q² + 0.5Q³ (₹).

• TFC = ₹100 (the term independent of Q); TVC = 20Q − 4Q² + 0.5Q³.
• MC = dTC/dQ = 20 − 8Q + 1.5Q². AVC = 20 − 4Q + 0.5Q².
• AVC minimum: dAVC/dQ = −4 + Q = 0 → Q = 4; AVC = 20 − 16 + 8 = ₹12. Check MC at Q = 4: 20 − 32 + 24 = ₹12 = AVC — MC cuts AVC at its minimum.
• MC minimum: dMC/dQ = −8 + 3Q = 0 → Q = 8/3 ≈ 2.67, MC = 20 − 21.33 + 10.67 ≈ ₹9.33.
• Shut-down price = minimum AVC = ₹12.

Example 4 — least-cost input mix. w = ₹200 per worker-day, r = ₹500 per machine-day. At the current mix MPL = 40 units, MPK = 80 units.

• MPL/w = 40/200 = 0.20 units per ₹; MPK/r = 80/500 = 0.16 units per ₹.
• Labour gives more output per rupee → hire more labour, use less capital. As L rises MPL falls and as K falls MPK rises, until the ratios are equal.
• Equilibrium requires MPL/MPK = w/r = 200/500 = 0.4; so at the optimum, if MPK = 100 then MPL = 40.

Example 5 — returns to scale and break-even. Q = 10 L^0.6 K^0.3.

• α + β = 0.9 < 1 → decreasing returns to scale. Doubling both inputs multiplies output by 2^0.9 ≈ 1.866, i.e. +86.6%, not +100%.
• Output elasticity of labour = 0.6: a 10% rise in L alone raises Q by about 6%.
• Break-even: a unit sells at ₹150 with AVC ₹90 and TFC ₹3,60,000 per month. Contribution = ₹60 → BEQ = 3,60,000/60 = 6,000 units/month; break-even sales = ₹9,00,000. To earn ₹1,20,000 profit: (3,60,000 + 1,20,000)/60 = 8,000 units.

Where it is used

• Manufacturing and FMCG (Tata Steel, Hindustan Unilever, Amul): plant-size decisions, standard costing, make-or-buy, and contribution analysis for product-mix choices.
• Regulated utilities: CERC and State Electricity Regulatory Commissions fix tariffs on fixed-plus-variable (capacity and energy) charges — a direct application of TFC/TVC separation; PNGRB does the same for gas pipelines.
• Competition policy: CCI uses average variable cost as the benchmark for predatory pricing under the Competition Act, 2002 (as amended in 2023).
• Banking and credit appraisal: project appraisal notes at SBI/NABARD compute break-even, capacity utilisation and DSCR from the same cost structure; MSME viability studies use break-even output.
• Consulting and analyst roles: cost-curve benchmarking, learning-curve pricing in electronics and solar module manufacturing, PLI-scheme scale justifications.
• Agriculture and policy: CACP computes A2, A2+FL and C2 costs of cultivation for MSP recommendations — C2 includes imputed rent and interest, a textbook implicit-cost calculation.

Interview questions

• Q: What is the difference between returns to a factor and returns to scale? A: Returns to a factor is short-run — one input varies against a fixed input; returns to scale is long-run — all inputs change in the same proportion.
• Q: Why is the AVC curve U-shaped? A: Because average product of the variable factor first rises and then falls (AVC = w/AP), so cost per unit falls then rises.
• Q: Why does MC cut AC at AC's minimum? A: AC falls while MC is below it and rises once MC is above it, so the two must be equal exactly where AC stops falling.
• Q: Does a rise in fixed cost change marginal cost? A: No — MC is the derivative of total variable cost; fixed cost shifts TC and AC up but leaves MC unchanged.
• Q: A factory is making a loss. When should it still keep running? A: When price at least covers average variable cost; it then recovers some fixed cost, so shutting down would mean a larger loss.
• Q: Why is sunk cost ignored in decisions? A: It is already spent and unrecoverable, so it is the same under every alternative and cannot change which option is best.
• Q: Accounting profit is ₹10 lakh — can economic profit be zero? A: Yes, if implicit costs (owner's salary, rent on own premises, interest on own capital) total ₹10 lakh; the firm is then earning only normal profit.
• Q: What does α + β = 1 mean in a Cobb-Douglas function? A: Constant returns to scale and linear homogeneity, so doubling all inputs exactly doubles output and factor payments exhaust the product.
• Q: Why is LAC called the planning curve and why is it flatter than SAC? A: It shows the cheapest plant for each output before any plant is built; since all inputs can adjust, cost rises less steeply with output.
• Q: If MPL/w > MPK/r, what should the firm do? A: Substitute labour for capital — labour yields more output per rupee — until the two ratios are equal.
• Q: Can a firm have increasing returns to scale and diminishing returns to labour at the same time? A: Yes — they refer to different experiments: all inputs scaled up versus one input increased against a fixed input.
• Q: Why do economists call normal profit a cost? A: It is the opportunity cost of the entrepreneur's own capital and effort — the minimum needed to keep resources in this use.

Common mistakes

• Confusing diminishing returns (short run, one factor) with decreasing returns to scale (long run, all factors).
• Drawing MC through the minimum of AC only, or showing AVC and AC bottoming out at the same output.
• Treating AFC as reaching zero, or forgetting that it is a rectangular hyperbola.
• Including fixed cost in marginal cost, or adding sunk cost into a make-or-buy or shut-down calculation.
• Using accounting profit as economic profit — omitting implicit costs and normal profit.
• Saying the firm shuts down whenever it makes a loss, instead of comparing price with minimum AVC.
• Mistaking the LAC as tangent to the minimum point of every SAC; it is tangent at SAC minima only at the LAC's own minimum.
• Reading α and β in Cobb-Douglas as returns to scale individually instead of summing them.

Related topics

Perfect competition and monopoly · Demand and supply · Factor pricing: wages, rent, interest and profit · Managerial economics · Oligopoly and game theory basics · Welfare economics and market failure