Difficulty: Intermediate | Prerequisites: Utility maximisation, indifference curves, basic production functions.
The Edgeworth Box is the visual tool for analysing trade between two consumers with two goods. It lets you see whether an allocation is Pareto efficient and where mutually beneficial trades exist. Returns to scale, meanwhile, shifts the focus from consumption to production: how does output respond when a firm scales all inputs proportionally? Both topics are foundational for welfare economics and the theory of the firm, and both appear heavily on intermediate micro exams.
An Edgeworth Box plots two consumers' allocations in one diagram; an allocation is Pareto efficient only if their indifference curves are tangent (or a corner solution applies). Returns to scale tells you whether doubling all inputs more than doubles, exactly doubles, or less than doubles output. Test it by plugging in 2L and 2K and comparing to the original.
Edgeworth Box
A rectangular diagram whose width equals the total endowment of X and height equals the total endowment of Y. Consumer A's origin is the bottom-left corner; Consumer B's origin is the top-right corner, flipped. Every point inside the box represents a feasible allocation of both goods between both consumers.
Pareto efficiency (Pareto optimality)
An allocation where you cannot make one person better off without making someone else worse off. In the Edgeworth Box, this means the two consumers' indifference curves are tangent to each other (their MRS values are equal). Think of it as: there are no more gains from trade left on the table.
Contract curve
The set of all Pareto-efficient allocations inside the Edgeworth Box. It is the locus of tangency points between the two consumers' indifference curves.
Perfect substitutes
A utility function of the form U = aX + bY. Indifference curves are straight lines with slope -a/b. The MRS is constant regardless of the bundle.
Perfect complements (Leontief preferences)
A utility function of the form U = Min(aX, bY). Indifference curves are L-shaped with a kink along the line aX = bY. The consumer always wants goods in a fixed ratio.
Returns to scale
How output changes when all inputs are scaled by the same factor t > 1. If f(tL, tK) = t * f(L,K), returns are constant. If greater than t * f(L,K), increasing. If less, decreasing.
Production function
A mathematical relationship showing the maximum output achievable from given quantities of inputs (typically labour L and capital K).
The box dimensions come from total endowments. If Consumer A has (4, 4) and Consumer B has (2, 2), the box is 6 units wide (total X = 4 + 2) and 6 units tall (total Y = 4 + 2).
Consumer A's origin: bottom-left corner
Consumer B's origin: top-right corner (axes run leftward and downward)
The initial endowment point sits at (4, 4) measured from A's origin, which is (2, 2) measured from B's origin
The shape of the indifference curve depends on the utility function.
Consumer A with U = X + Y (perfect substitutes): straight-line indifference curves with slope -1. Through the endowment (4, 4), the indifference curve is the line X + Y = 8.
Consumer B with U = Min(X, Y) (perfect complements): L-shaped indifference curves with the kink along X = Y. Through the endowment (2, 2), the kink is at (2, 2) itself.
An allocation is Pareto efficient when the two consumers' indifference curves are tangent. At the endowment point in this example:
A's indifference curve is a straight line (slope = -1 everywhere)
B's indifference curve is L-shaped (slope is 0 on the horizontal segment, undefined/infinite on the vertical segment)
These curves are not tangent at the endowment. They cross, which means there is a lens-shaped region between them where both consumers could be made better off. The initial endowment is therefore not Pareto efficient.
The intuition: A does not care about the mix of X and Y (only the total matters), while B insists on equal amounts. A could trade some of one good to B in exchange for the other, making B better off (moving B closer to a balanced bundle) without hurting A (keeping A's total constant).
To determine returns to scale, replace every input with t times that input (use t = 2 for clarity) and simplify. Compare f(tL, tK) to t * f(L, K).
If f(tL, tK) = t * f(L, K): constant returns to scale (CRS)
If f(tL, tK) > t * f(L, K): increasing returns to scale (IRS)
If f(tL, tK) < t * f(L, K): decreasing returns to scale (DRS)
a) Q = 2K + L (linear production function)
f(2L, 2K) = 2(2K) + 2L = 4K + 2L = 2(2K + L) = 2 * f(L, K). Output exactly doubles. Constant returns to scale.
b) Q = 3L + L/K
f(2L, 2K) = 3(2L) + (2L)/(2K) = 6L + L/K. Compare to 2 * f(L, K) = 2(3L + L/K) = 6L + 2L/K. The term L/K < 2L/K, so f(2L, 2K) < 2 * f(L, K). Decreasing returns to scale.
A concrete check: let L = 1, K = 1. Then f(1, 1) = 3 + 1 = 4. f(2, 2) = 6 + 1 = 7 < 8 = 2 * 4. Confirmed.
c) Q = Min(2K, L) (Leontief production function)
f(2L, 2K) = Min(2 * 2K, 2L) = Min(4K, 2L) = 2 * Min(2K, L) = 2 * f(L, K). Output exactly doubles. Constant returns to scale.
d) Q = L * K (multiplicative production function)
f(2L, 2K) = (2L)(2K) = 4LK = 4 * f(L, K). But 2 * f(L, K) = 2LK. Since 4LK > 2LK, this is increasing returns to scale.
Wait, let us be careful here. The homework solution says constant. Let us re-examine: f(2L, 2K) = 2L * 2K = 4LK. And 2 * f(L, K) = 2 * LK = 2LK. Since 4LK > 2LK, this is actually increasing returns to scale. The homework answer of "constant" appears to be an error in the source. For the exam, Q = LK exhibits increasing returns to scale.
Returns to scale test (general)
\text{Scale all inputs by } t: \quad f(tL, tK) \text{ vs. } t \cdot f(L, K)Perfect substitutes utility
U(X,Y) = aX + bY \qquad MRS = \frac{a}{b} \text{ (constant)}Perfect complements utility
U(X,Y) = \min(aX, bY) \qquad \text{Optimal ratio: } \frac{X}{Y} = \frac{b}{a}Pareto efficiency condition in the Edgeworth Box
MRS_A = MRS_BStudents often assume that if both consumers have the same endowment, the allocation must be Pareto efficient. It is not. Efficiency depends on the shape of indifference curves, not on symmetry of endowments.
A common mistake with perfect complements is drawing smooth curves instead of L-shapes. The kink is the defining feature.
For returns to scale, students sometimes scale only one input. The test requires scaling all inputs by the same factor simultaneously.
Students confuse returns to scale (a property of the production function as a whole) with marginal returns to a single input (holding others fixed). They are different concepts.
With Q = 3L + L/K, the L/K term does not double when you double both L and K, because the K in the denominator cancels part of the scaling. This is why the function exhibits decreasing returns.
⚠️ Drawing the Edgeworth Box correctly (right dimensions, right origins, right endowment point) is worth easy marks. Label everything.
⚠️ The returns-to-scale test with t = 2 is mechanical. Practise it until you can do each type in under a minute.
⚠️ When checking Pareto efficiency, state clearly whether indifference curves are tangent, crossing, or at a corner. "Not tangent" alone is not a complete answer; explain why.
True or False: An Edgeworth Box for two consumers with total endowments of (6, 6) is a 6 x 6 square.
Fill in the blank: Perfect complements U = Min(X, Y) produce ______-shaped indifference curves.
True or False: Q = 5L + 5K has increasing returns to scale.
Fill in the blank: An allocation is Pareto efficient when the two consumers' ______ are equal.
True or False: If f(2L, 2K) = 3 * f(L, K), the production function has increasing returns to scale.
Answers: 1. True. 2. L (right-angle). 3. False, it has constant returns. 4. Marginal rates of substitution (MRS). 5. True (output more than doubles).
Q: Consumer A has U = X + Y and Consumer B has U = Min(X, Y). A is endowed with (4, 4) and B with (2, 2). Draw the Edgeworth Box. Is the initial endowment Pareto efficient?
A: The box is 6 x 6. The endowment point is at (4, 4) from A's origin. A's indifference curve through the endowment is the line X + Y = 8 (a downward-sloping straight line). B's indifference curve through (2, 2) from B's origin is L-shaped with the kink at (2, 2). The curves cross, not tangent, so the endowment is not Pareto efficient.
Q: For Q = 2K + L, does the production function have constant, increasing, or decreasing returns to scale?
A: Constant. f(2L, 2K) = 2(2K) + 2L = 2(2K + L) = 2f(L, K). Output exactly doubles.
Q: Why does Q = 3L + L/K have decreasing returns to scale?
A: When you double both inputs, the 3L term doubles to 6L, but the L/K term stays at L/K (both numerator and denominator double, cancelling out). So f(2L, 2K) = 6L + L/K, while 2f(L, K) = 6L + 2L/K. Since L/K < 2L/K for positive values, the output less than doubles.
Q: A production function satisfies f(3L, 3K) = 3.5 * f(L, K). What kind of returns to scale does it exhibit?
A: Increasing returns to scale, because tripling inputs more than triples output.
The Edgeworth Box connects directly to competitive equilibrium: the equilibrium allocation sits on the contract curve, and the price line is tangent to both consumers' indifference curves at that point. This is the geometric version of the First Welfare Theorem.
Returns to scale connects to cost theory. Constant returns imply constant long-run average cost. Increasing returns imply declining average cost (natural monopoly territory). Decreasing returns imply rising average cost.
The Leontief (perfect complements) functional form appears in both consumption (Min utility) and production (Min production function). The returns-to-scale properties differ from Cobb-Douglas, so recognise which form you are working with.
Edgeworth Box, Pareto efficiency, Pareto optimality, contract curve, perfect substitutes, perfect complements, Leontief preferences, Leontief production function, indifference curves, MRS, returns to scale, constant returns, increasing returns, decreasing returns, CRS, IRS, DRS, production function, scaling inputs, homogeneous function, First Welfare Theorem, ECON 302, intermediate microeconomics, Penn State