Think of a farmer with a fixed plot of land (say, one acre). On day one, he works alone and produces 10 kg of wheat. He brings in a second worker — output jumps to 22 kg. A third worker — 30 kg. A fourth — 35 kg. A fifth — 38 kg. A sixth — 39 kg. A seventh? Maybe still 39 kg, or even less if they start getting in each other's way.
That story — where adding more of one input to a fixed input eventually yields smaller and smaller additions to output — is the entire intuition behind the Theory of Production and Cost.
Production theory asks: how do inputs transform into outputs? Cost theory asks: what does it cost to produce each unit at different output levels? These two questions are inseparable in microeconomics because your cost structure is a direct reflection of your production technology.
Production Function is the technical relationship between inputs (capital, labour, land, etc.) and the maximum output achievable from them. It can be written as Q = f(K, L) where Q is output, K is capital, and L is labour.
Two time horizons matter here:
Here is the analogy that makes this stick: imagine your laptop's RAM is fixed (short run). You can open more browser tabs (more variable input = labour analogy), but at some point, adding another tab slows everything down. In the long run, you can also upgrade the RAM (all inputs variable), and now scaling up everything proportionally — that is returns to scale.
Cost theory mirrors this. Short-run costs split into fixed costs (rent, loan repayments — you pay these regardless of output) and variable costs (raw materials, wages — they move with output). Long-run costs are all variable.
When you add successive units of a variable input (labour) to a fixed input (capital or land), output first rises at an increasing rate, then at a decreasing rate, and eventually may fall. This generates three distinct phases of the Total Product (TP) curve:
| Phase | Marginal Product (MP) | Average Product (AP) | Description | |---|---|---|---| | I — Increasing Returns | Rising | Rising | Specialisation gains; MP > AP | | II — Diminishing Returns | Falling but positive | Falling | MP < AP; TP still rising | | III — Negative Returns | Negative | Falling | TP falls; too many workers |
Marginal Product (MP) = change in total output from adding one more unit of the variable input.
The law says MP will eventually decrease — the word "eventually" is critical. MP may initially rise (Phase I), but it must eventually fall (Phase II and III). SSC CGL questions often test whether you catch the word "eventually" — the law does not say MP falls immediately.
When ALL inputs are increased proportionally:
The key distinction to drill: Diminishing Marginal Returns is a short-run concept (one input fixed, one varied). Decreasing Returns to Scale is a long-run concept (all inputs scaled proportionally). SSC CGL regularly mixes these up in answer options — this distinction is a perennial trap.
An isoquant (equal-quantity curve) shows all combinations of capital and labour that produce the same output level. It is the production-side equivalent of the indifference curve from consumer theory.
Key isoquant shapes reveal the nature of inputs:
ΔK/ΔL.Break total cost into fixed and variable components:
TC = TFC + TVCAFC = TFC / Q — always falls as Q rises (rectangular hyperbola shape)AVC = TVC / Q — U-shapedATC (or AC) = TC / Q = AFC + AVC — U-shaped, lies above AVCMC = ΔTC / ΔQ — also U-shaped; cuts both AVC and ATC at their respective minimum pointsLook — this is one of the most frequently tested relationships in SSC CGL Economics. Here is the logic, not just the rule:
Think of your exam marks. Your "average score" across 10 tests is 60. If your 11th test score (the "marginal" score) is 50 — below your average — your average falls. If the 11th score is 70 — above your average — your average rises. If it equals 60, the average stays the same.
Apply this to cost curves:
MC < AC → AC is fallingMC > AC → AC is risingMC = AC → AC is at its minimum (MC intersects AC at AC's lowest point)The same logic holds for the MC-AVC relationship: MC cuts AVC at AVC's minimum, which occurs at a lower output level than where MC cuts AC.
If TC = 0.5Q³ − 10Q² + 100Q + 200:
VC = TC − FC = 0.5Q³ − 10Q² + 100QAVC = VC/Q = 0.5Q² − 10Q + 100d(AVC)/dQ = Q − 10 = 0 → Q = 10d²(AVC)/dQ² = 1 > 0 — confirms minimumThis calculus-based approach is tested in the more challenging CGL sets. Even if you do not use calculus, recognise that AVC is quadratic in Q and behaves like a U-shaped parabola.
All inputs change → Returns to Scale (long run). One input changes, others fixed → Diminishing Marginal Returns (short run). When you see a question, count the inputs changing. If the question says "doubles ALL inputs," it is returns to scale. If it says "adds more labour to a fixed factory," it is the law of variable proportions. Standard confusion time: 30 seconds re-reading. With this pattern: 5 seconds — done.
When a new exam score (MC) is below your current average (AC), average drops. When above, average rises. Direct substitution into any question: read whether MC is above or below AC, then state whether AC is rising or falling. Standard derivation from memory: 4 steps. With this analogy: 1 step — read the inequality and state the direction. Eliminates 3 wrong options in under 8 seconds.
AFC = TFC / Q. TFC never changes (it is fixed by definition). So as Q increases, you divide the same fixed number by a larger denominator every time — AFC must always fall and never rises. Questions that suggest AFC rises or stays constant are automatically wrong. Elimination of 2-3 wrong options: under 5 seconds, no calculation needed.
Map shape to substitutability: Straight line → Perfect Substitutes (constant MRTS). L-shape (right angle) → Perfect Complements (zero substitutability). Normal bow → Some substitutability (diminishing MRTS). If a question gives you the shape and asks for the input relationship, substitute the shape directly into this three-point map. Standard reasoning time: 25 seconds. With the map memorised: 8 seconds.
For any numerical cost question, always calculate TVC = TC − TFC before dividing. Students who divide TC by Q get ATC, not AVC — the most common numerical error in this chapter. Two-step sequence: (1) TVC = TC − TFC, (2) AVC = TVC / Q. This eliminates the ATC option that always appears as a distractor. Standard error rate on this type: high. With the subtract-first rule: near-zero error in 10 seconds.
When you hit a production/cost question in the exam hall, run this quick decision tree:
Step 1 — Identify the time horizon. Does the question mention "fixed" input or "short run"? → Law of Variable Proportions or short-run cost curves. Does it say "all inputs" or "long run"? → Returns to Scale or long-run costs.
Step 2 — Identify what is asked. Shape of a curve? Use the standard shapes (AFC = hyperbola, AVC/ATC/MC = U-shaped). A formula? AFC = TFC/Q, AVC = TVC/Q, ATC = TC/Q, MC = ΔTC/ΔQ. A relationship? Use the MC-AC exam-score analogy.
Step 3 — For numericals. Always find TVC = TC − TFC first. Then divide by Q for AVC. Never skip the subtraction step.
Step 4 — Returns to Scale test. Compare percentage change in output to percentage change in inputs. Output% > Input% → IRS. Output% = Input% → CRS. Output% < Input% → DRS.
Most questions resolve within 20-30 seconds using steps 1 and 2 alone.
Why this question: This is the most fundamental definition question in the chapter — tests whether you know "eventually" is the operative word in the law.
Solving path: The law does not say MP becomes negative immediately (option C is a trap). It says MP eventually decreases. Phase I may even show rising MP. The answer is "Decrease" — option D. Time: 10 seconds.
Why this question: AFC formula is tested almost every year across SSC tiers. The rectangular hyperbola description is the visual anchor.
Solving path: AFC = TFC / Q. The only option with TFC in the numerator is option D. Eliminate: option A is AVC, option B is MC, option C is ATC. Time: 8 seconds by elimination.
Why this question: This is the classic DRS vs DMR trap — the most common conceptual error in this chapter.
Solving path: "Doubles ALL inputs" → this is a Returns to Scale question, not diminishing marginal returns. Output more than doubles → output increases proportionally more than inputs → Increasing Returns to Scale. Option D. Time: 15 seconds.
Why this question: Standard numerical — tests the TVC extraction step before computing AVC.
Solving path: TVC = TC − TFC = 8,000 − 2,000 = 6,000. AVC = 6,000 / 100 = 60. Answer: ₹60. The distractor ₹80 is ATC (8,000/100) — exactly the error you get if you skip the subtraction step. Time: 12 seconds.
Why this question: The MC-AC relationship is a perennial SSC CGL favourite — tested in multiple forms.
Solving path: Use the exam-score analogy. When MC < AC → AC falls. When MC > AC → AC rises. Option D matches this exactly. Option A reverses the relationship — the classic distractor. Time: 8 seconds.
Why this question: Tests the calculus-based approach to finding minimum AVC — a harder variant seen in recent CGL papers.
Solving path: FC = 200 (constant term). VC = 0.5Q³ − 10Q² + 100Q. AVC = VC/Q = 0.5Q² − 10Q + 100. Minimise: d(AVC)/dQ = Q − 10 = 0 → Q = 10. Answer: option B. Even without calculus, recognise AVC is a parabola in Q; the vertex of 0.5Q² − 10Q + 100 occurs at Q = 10/(2×0.5) = 10. Time: 30 seconds.
Why this question: Tests the numerical definition of DRS — percentage comparison is the key.
Solving path: Inputs doubled = 100% increase. Output increases by 80% — less than 100%. Output% < Input% → Decreasing Returns to Scale. Option A. Time: 10 seconds.
Why this question: Isoquant shapes are regularly tested and often confused.
Solving path: Straight line + negative slope + constant slope = constant MRTS = perfect substitutes. L-shaped = perfect complements (option B, wrong). Bowed isoquant = diminishing MRTS (option A, wrong). Answer: option C. Time: 10 seconds using the shape-substitutability map.
Confusing Diminishing Marginal Returns with Decreasing Returns to Scale. These are different concepts operating in different time horizons. Diminishing marginal returns: one input fixed, one varied, short run. Decreasing returns to scale: all inputs scaled proportionally, long run. When an option offers both as possible answers, check whether the question says "all inputs" or "fixed input."
Assuming MP falls from the very first unit. The law says MP eventually decreases, not immediately. In Phase I, MP may actually rise. Questions that say "MP decreases from the first unit" are describing a special case, not the general law.
Calculating AVC as TC/Q instead of TVC/Q. This gives you ATC, not AVC. Always compute TVC = TC − TFC first, then divide by Q. The trap answer (ATC value) always appears in the options.
Thinking AFC can rise or remain constant. AFC = TFC/Q. TFC is a constant. Dividing a constant by an increasing Q gives a continuously decreasing value. AFC never rises, never stays flat — it always falls. Any option suggesting otherwise is wrong by definition.
Placing the MC-ATC intersection anywhere other than ATC's minimum. MC must cut ATC exactly at ATC's lowest point — not to the left, not to the right. The same applies to MC cutting AVC at AVC's minimum, which occurs at a lower output level than the ATC minimum.
Misreading isoquant shapes. A straight-line isoquant means perfect substitutes (constant MRTS). L-shaped means perfect complements (zero substitutability). Students frequently swap these. Anchor: a straight line allows you to slide anywhere along it (substitution is free) — perfect substitutes. An L-shape forces you to stay at the corner (no substitution possible) — perfect complements.